TErrestrial Trunked RAdio - TETRA: A Global Security Tool (Signals and Communication Technology)

TErrestrial Trunked RAdio - TETRA Peter Stavroulakis TErrestrial Trunked RAdio - TETRA A Global Security Tool With 1...

Author: Peter Stavroulakis

478 downloads 1159 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

TErrestrial Trunked RAdio - TETRA

Peter Stavroulakis

TErrestrial Trunked RAdio - TETRA A Global Security Tool With 124 Figures and 22 Tables

123

E d itor Prof. Peter Stavroulakis Technical University C rete Aghiou Markou 731 32 C hania, C rete Greece E-mail: [email protected]

Library of Congress Control Number: 2007926105

ISBN 978-3-540-71190-2 Springer Berlin Heidelberg New York This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable for prosecution under the German Copyright Law. Springer is a part of Springer Science+Business Media springer.com © Springer-Verlag Berlin Heidelberg 2007 The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Typesetting: by the editors Production: Integra Software Services Pvt. Ltd., India Cover design: wmxDesign GmbH, Heidelberg

Printed on acid-free paper

42/3100/Integra

5 4 3 2 1 0

Dedication

This book is dedicated in memory of my sister Mary whose affection, support and guidance during the early years of my University studies affected my career in a fundamental way.

Acknowledgements

I feel indebted to the contributors of this book whose diligent work and unique contributions made this book possible. Special thanks are due to my collaborator Anastasios M. Haddad who spent countless hours to put the material in the format required by the publisher.

Preface

Following the development of wireless communications starting from satellites in the late 60’s to the wireless cellular in the 80’s and now the Private Mobile Radio (PMR) systems, it is obvious that each time every new technology achieved a good technical solution satisfying a special need at an acceptable price until another need would come up that needed special consideration. Of course the technology evolves at a different pace at different time periods through the centuries but one thing has to be true. Special needs always existed, but remained unsatisfied until technology could be used to offer an acceptable solution at a reasonable price. The last, at least five years, the public safety issue, became of major importance. Many types of Public/Private Mobile systems have appeared as we show in chapters 2 and 3, which seemed to satisfy the need at that time. The European Telecommunications Standards Institute (ETSI), foresaw that Terrestrial Trunking Radio - TETRA systems are good candidates to satisfy this need even on an international level. With the encouragement and support of the TETRA MOU, standards were in a continuous evolution and TETRA has become the tool to design any type of public security system. In this book we show the TETRA can be improved greatly and these improvements eventually will be part of future standards. The areas examined include channel assignment and multiple access techniques, video transmission, WLAN integration and the establishment of multiple wireless mesh networks. Since the requirements for these networks is security we show that innovative techniques such as those based on chaotic signals can be used to maximize security. It is believed that this book will become a reference point for many new researchers whose ambition is to find a general solution to the modern problems within the context of public safety.

List of Contributors

Peter Stavroulakis Technical University of Crete Crete, Greece Email: [email protected] Stavros Kotsopoulos Wireless Telecommunication Laboratory Department of Electrical and Computer Engineering University of Patras, Patras, Greece Email: [email protected] Dr Konstantinos Ioannou Wireless Telecommunication Laboratory Department of Electrical and Computer Engineering University of Patras, Patras, Greece Email: [email protected] Dr John Panoutsopoulos, Michail Tsagkaropoulos Wireless Telecommunication Laboratory Department of Electrical and Computer Engineering University of Patras, Patras, Greece Email: [email protected] [email protected] Ilias Politis Wireless Telecommunication Laboratory Department of Electrical and Computer Engineering University of Patras, Patras, Greece Email: [email protected]

X II List of Contributors

Apostolis Salkitzis Motorola, 32 Kiffisias Ave., Athens, Greece 15125 E-mail: [email protected] Dimitris Axiotis Telecommunications Laboratory School of Electrical and Computer Engineering National Technical University of Athens E-mail: [email protected] Pau Plans, Carles Gomez, Josep Lluis Ferrer and Josep Paradells Technical University of Catalonia (UPC) Wireless Networks Group (WNG) E-mail: [email protected]

Table of Contents

1 Introduction ............................................................................................. 1 1.1 Why TETRA .................................................................................. 1 References.................................................................................................... 4 2 Modern Security Requirements in Private Mobile Communications Systems........................................................................ 5 2.1 Introduction .................................................................................... 5 2.2 PMR Systems [1]............................................................................ 6 2.2.1 PMR Configurations ............................................................. 6 2.2.2 Comparison Between PMR and Cellular [2]....................... 11 2.2.3 PMR Standards [1].............................................................. 14 2.3 PMR Limitations [4] .................................................................... 28 2.3.1 Edge of Coverage Voice Quality ........................................ 28 2.3.2 Requirements of PMR Services .......................................... 33 2.3.3 Interoperability [6] .............................................................. 37 References ............................................................................................. 42 3 TETRA Providing an Acceptable Security System Solution ................ 43 3.1 Introduction .................................................................................. 43 3.2 Hierarchical analysis .................................................................... 44 3.2.1 Air interface specifications.................................................. 44 3.2.2 GSM ASCI .......................................................................... 45 3.2.3 Enhanced Multi-Level Precedence and Pre-emption service (eMLPP).................................................................. 45 3.2.4 Voice Group Call Service (VGCS) ..................................... 46 3.2.5 Voice Broadcast Service (VBS).......................................... 47 3.3 TETRA ......................................................................................... 47 3.3.1 Comparison of specified features ........................................ 48 3.3.2 Technical analysis ............................................................... 49 References ............................................................................................. 66 4 Channel Assignment and Multiple Access in Trunking Radio Systems [1].............................................................. 67 4.1 Channel Assignment Techniques [1]............................................ 67 4.1.1 Introduction ......................................................................... 67 4.1.2 Channel Allocation Schemes .............................................. 68

X IV Table of Contents

4.2

Channel Assignment Optimization............................................... 80 4.2.1 Introduction ......................................................................... 80 4.2.2 Model Formulation.............................................................. 80 4.2.3 One Layer Architecture using Erlang Model ...................... 82 4.2.4 Channel Assignment Scheme based on a Three Layer Architecture......................................................................... 84 4.2.5 Comparison of One layer with Three Layer Architecture......................................................................... 90 4.3 Multiple Access Techniques....................................................... 102 4.3.1 CDMA Techniques in TETRA systems .......................... 102 References ........................................................................................... 126 5 Video Transmission over TETRA ....................................................... 133 5.1 Introduction ................................................................................ 133 5.2 Evolution of Public Safety Mobile Networks............................. 134 5.2.1 Evolving Data services for public safety........................... 135 5.2.2 The TETRA solution to PSDR communication environment............................................. 136 5.2.3 The Market Considerations ............................................... 138 5.2.4 TETRA Enhanced Data Service-TEDS ............................ 139 5.3 Overview of DATA Transmission over TETRA ....................... 141 5.3.1 TETRA (V+D) Technical Characteristics......................... 141 5.3.2 TETRA Network Services ................................................ 147 5.3.3 High Speed Data service provisioning .............................. 149 5.4 Video Encoding Techniques....................................................... 151 5.4.1 Background ....................................................................... 151 5.4.2 Compression standards overview...................................... 153 5.4.3 Encrypted Video over TETRA.......................................... 170 5.5 Performance Analysis of video broadcasting over TETRA ....... 174 5.5.1 Performance Evaluation .................................................... 175 5.5.3 Video Quality Measurements............................................ 178 5.6 Vision for Future Public Safety Communication Systems ......... 181 5.6.1 Future Trends .................................................................... 181 5.6.2 All-IP convergence............................................................ 182 5.6.3 TETRA – TEDS interoperability ...................................... 183 5.6.4 TETRA over IP ................................................................. 183 5.6.5 Integrated TETRA-WLAN system ................................... 184 5.7 Conclusions ................................................................................ 186 References ........................................................................................... 188

Table of Contents X V

6 TETRA as a Gateway to Other Wireless Systems.............................. 191 6.1 Introduction ................................................................................ 191 6.2 TETRA Air Interface: Logical and Physical Channels .............. 192 6.2.1 Logical Channels............................................................... 193 6.2.2 Physical channels .............................................................. 194 6.3 TETRA Packet Data Transmission ............................................ 195 6.3.1 Packet Data transmission and reception procedures ......... 198 6.3.2 TETRA IP user authentication .......................................... 202 6.4 SNDCP states and state transitions............................................. 205 6.5 UDP versus TCP on top of TETRA IP layer.............................. 211 6.6 TETRA Packet Data modems .................................................... 213 6.6.1 Types of Packet-data Mobile Stations............................... 214 6.7 TETRA and WLAN Integration for Improving Packet-Data Transmission Capabilities .......................................................... 216 6.7.1 Integrated WLAN/TETRA System Overview .................. 220 6.8 System Architecture ................................................................... 223 6.8.1 Architecture Elements and Interfaces ............................... 223 6.8.2 Protocol Architecture ........................................................ 225 6.8.3 Packet Structure ................................................................ 227 6.8.4 WLAN Association and TETRA Location Update Procedure .............................................................. 228 6.8.5 Group Call Initiation and Participation ............................. 230 6.9 Conclusions ................................................................................ 231 References ........................................................................................... 233 7 TETRA as a Building block to WMNs................................................ 235 7.1 Introduction ................................................................................ 235 7.1.1 Requirements..................................................................... 239 7.1.2 Discussion ......................................................................... 244 7.2 Wireless Mesh Networks............................................................ 245 7.2.1 Definition and classification of WMNs ............................ 245 7.2.2 MANET routing protocols ................................................ 246 7.2.3 Influence of routing protocols on network performance... 253 7.2.4 Multicast in WMNs ........................................................... 259 7.3 TETRA DMO............................................................................. 263 7.3.1 DMO overview.................................................................. 263 7.4 TETRA Release 2....................................................................... 273

X VI Table of Contents

7.5

TETRA extensions for building WMNs..................................... 275 7.5.1 Routing capabilities........................................................... 277 7.5.2 Wireless Interface.............................................................. 283 7.5.3 Overview of network performance figures ....................... 287 7.6 Conclusion.................................................................................. 293 References ........................................................................................... 295 Appendix.................................................................................................. 299

1 Introduction

Peter Stavroulakis

1.1 Why TETRA If someone has been following the recent (last five years) International events both in the sphere of technical developments but also in the International politics, he would have noticed that there is a general outcry for the development of secure telecommunication systems covering all aspects of wireless communications. This situation has become much more serious and demanding because of the fact that wireless technology evolution is accelerating to satisfy the ever increasing market demands for more mobility to all kinds of telecommunications services whether they involve voice, IP, Video, WLANS, Ad-Hoc, Mesh, Peer- to-Peer and Sensor Networks to name a few. This newly created demand, a large part of which comes from the public and private safety sectors such as police, fire brigades, ambulances (telemedicine) and the private sector such as the trucking transport businesses, airport safety authorities e.t.c, is being satisfied by the Public/Private Mobile Radio(PMR) systems. These systems have, on one hand, the general characteristics of cellular systems but, on the other hand possess such unique security characteristics such as end to end encryption, field control by a dispatcher type of capabilities as well as trunking(switching) capability and thus create autonomous telecommunications systems and distinct from cellular. In this book, we are examining this evolution and we are showing that TETRA is the best candidate to satisfy all above requirements whether they are technological, security or service oriented. The reader who is not familiar with the fundamental aspects of PMR or TETRA is referred to [1 and 2]. Since TETRA was developed mainly for voice communications,

2

1 Introduction

the European Telecommunications Standards Institute (ETSI) which is responsible for developing its standards at least for Europe, is continuously updating its them in order to meet new requirements. See chapter 7. We, in this book, examine new ways and go one step further to prove that further improvements and innovative techniques, which we hope one day will become TETRA standards, can make TETRA the building block for future security networks for universal use. As a matter of fact we propose, in the Appendix, a general scheme which is based on the results presented in the chapters 2-7 and on references 2 and 3 that TETRA not only could become the building block for integrating WLANS and Ad-Hoc networks into unified networks for providing general wireless services but will also serve as a global security tool through the use of the proposed chaotic based encryption and/or modulation techniques.[3, 4]. We show in Chapter 2 the basic features of the class of systems to which TETRA belongs, their basic configurations, the different technologies used and the problems that present in their usage as security tools in the Private Mobile Radio (PMR) communications field. Actually, even though these systems have been designed as the security alternatives of their public equivalents such as the GSM, still they have limitations which are pointed out. As the best candidate to satisfy modern security requirements, we present TETRA. It is shown, by identifying the elements on which a comparison of the requirements with its special features of the evolving standards and the improvement that are possible for TETRA, that TETRA can play a major role in the next generations of PMR systems. This comparison takes place in chapter 3 where the differences between TETRA and a GSM solution are analysed in three dimensions in the hope to exhibit the clear and perhaps unique advantages of TETRA for security applications as will be shown in the subsequent chapters. The first dimension compares the applicable ETSI specifications and points out which functions are available according to the standards. Proprietary solutions will not be discussed on that level. The second dimension, a technical analysis, discusses how the end users and the operators perceive the differences between the network-solutions. Since it is possible technically to provide TETRA capabilities using GSM, an economic analysis focusing on the cost of the two alternative solutions, including capital and operational costs for the network infrastructure and end-user terminals which constitutes the third dimension, will be touched upon only briefly because the applications of secure systems do not depend as much on economic terms but on their technical feasibility and the existence of international standards and on their ability to satisfy certain predefined requirements. We go one step further, in chapter 4, to show that the TETRA system can be improved to become an unique tool for security. First area of enhancement is the

1.1 Why TETRA

3

channel assignment methodology. We first provide an overview of different channel assignment algorithms as they relate to TETRA Networks and compare them in terms of performance, flexibility, and complexity. We first start by giving an overview of the channel assignment problem in a cellular environment and we discuss the general idea behind major channel allocation schemes. Then we proceed to discuss different channel allocation schemes within each category and we follow with the development of an optimization technique in channel assignment. Finally we present multiple access techniques in section 4.3. In chapter 5 we present the scheme by which we can transmit securely VIDEO through TETRA. Within the context of this chapter, we monitor the QoS regarding MPEG-4 video streaming traffic delivery, in terms of both packet loss and perceived image quality, over TEDS networks. The rest of the chapter is organized as follows. In Section 5.2, we attempt to follow the evolution in public safety mobile networks and the role of TETRA networks. The need for more complex context in the information exchange, guaranteed quality of service and secure, flexible and scalable infrastructures led to the standardization of TETRA and TEDS systems. Section 5.3, includes an extensive overview of data transmission over TETRA, followed by an insight on the evolutions incorporated in TEDS standard that allowed video and high data rates support. A detailed analysis of the current video encoding techniques and error concealment methods are contained in Section 5.4. Particular interest is given to MPEG4 encoding standard as it enables higher video compression rates, thus making it an ideal solution for video traffic over TETRA and TEDS networks. The main two techniques of video encryption are described in this section as well. In the following Section 5.5, we provide a performance analysis of video transmission over TEDS network. In this chapter 6, we present an overview of packet data transmission over Terrestrial Trunked Radio (TETRA) release 1 networks as well as a solution for integrating TETRA with WLANs as a way to improve the packet data transmission capabilities. We first give a brief overview of the TETRA air interface and the available logical and physical channels. We then present various aspects of packet data transmission over TETRA, where we conclude that TETRA release 1 cannot provide the means to support demanding IP-based applications, mainly due to bandwidth and QoS constrains. Motivated from this conclusion, we then present a solution for integrating TETRA with Wireless Local Area Networks (WLANs) and thus realizing hybrid broadband networks suitable to support the next generation of public safety communication systems. The specified solution allows TETRA terminals to interface to the TETRA Switching and Management Infrastructure (SwMI) over a broadband WLAN radio access

4

1 Introduction

network, instead of the conventional narrowband TETRA radio network. These terminals are fully interoperable with conventional TETRA terminals and can employ all TETRA services, including group calls, short data messaging, packet data, etc. Possible extensions of the TErrestrial Trunked RAdio (TETRA) system with the purpose of building a mesh network are analyzed in chapter 7. The main objective is to provide the services, already offered by PMR systems, everywhere and at any moment, taking then profit of the natural advantages that a mesh network presents. We evaluate extensions which make use of already available functionality of TETRA, such as client, relay and gateway functions, in order to minimize the changes that should be made to the standard and make the adaptation as simple as possible. Finally in the Appendix we show that TETRA can be used as a building block for universal super-secure systems using chaotic techniques.

References 1. J. Dunlop et.al, Digital Mobile Communications and the TETRA System, John Wiley 1999. 2. Doug Gray, TETRA : The Advocate’s Handbook, Pryntya, 2003 3. Peter Stavroulakis, Chaos Applications in Telecommunications, Taylor and Francis 2006 4. Peter Stavroulakis, Secure Telecommunication Systems based on Chaotic and Interference Reduction Techniques, Pending International Patent PCT/GR000038

2 Modern Security Requirements in Private Mobile Communications Systems

Peter Stavroulakis, Kostas Ioannou

2.1 Introduction Having discussed the main objectives of the book in the previous chapter, we shall now embark on showing the basic features of the class of systems to which TETRA belongs, their basic configurations, the different technologies used and the problems that present in their usage as security tools in the Private Mobile Radio (PMR) communications field. Actually, even though these systems have been designed as the security alternatives of their public equivalents such as the GSM, still they have limitations which are pointed out. As the best candidate to satisfy modern security requirements, we present TETRA. It is shown, by identifying the elements on which a comparison of the requirements with its special features of the evolving standards and the improvement that are possible for TETRA, that TETRA can play a major role in the next generations of PMR systems. An area which requires special attention is the interoperability functional requirements based on both technical and operational issues The superiority of TETRA is then proven in chapter 3 comparing TETRA with its closest competitor which is GSM. This superiority becomes then the basis for using TETRA as the building block in many applications that involve implementations of secure integrated designs.

6

2 Modern Security Requirements

2.2 PMR Systems [1]

2.2.1 PMR Configurations The simplest PMR configuration is point-to-point direct terminal communication. Such a system has no infrastructure, and in most cases all terminals within range receive messages as shown in figure 2.1. It is possible, however, to conduct private conversations through the use of signaling tones or messages which mute terminals which the message is not intended for so that it is not heard. Either a single common frequency is used, or different frequencies can be used for different call groups. Communication is only possible between terminals when they are in range of each other, and given the power limitations on battery operated portable devices, this may be a significant restriction.

Fig. 2.1. Simple direct mode PMR configuration

One of the most common PMR configurations is the dispatch operation. At least two channels are used, one for uplink communications between terminals and the base station, and one for the downlink to the terminals. Messages from the dispatcher on the downlink can be received by all terminals (although again individual addressing is possible), whereas messages from the terminals can only be received by the dispatcher. Mobile to mobile communication is possible via the dispatcher. Links with the public switched telephone or data networks are possible, again via the dispatcher as shown in figure 2.2.

2.2 PMR Systems[1]

7

Fig. 2.2. Dispatch mode PMR configuration

A number of refinements to this basic system are possible. If extended coverage is required, but central dispatch or PSTN network access is not necessary, the base station can be connected as a repeater. This is called “talkthrough” mode where any uplink messages are retransmitted on the downlink, effectively extending the range of mobiles to that of the base station. In Figure 2.3, the transmission from mobile 1 is received by mobiles 2 and 3, even though they would not have been in range if the message had been transmitted directly.

8


Fig. 2.3. Talkthrough repeater operation

Different organizations can share repeaters (so-called “community base stations” or “community repeaters”) if the different users have signalling to identify their messages. The signalling is retransmitted by the base station, so that users in other groups are muted and privacy maintained. Since users in groups do not hear all the messages it is necessary to keep usage low to ensure access. Such systems therefore include time outs to ensure that users do not hog a channel. A better option, although one which requires more complexity, is trunked operation. In this case, several channels are available, pooled between different PMR operators. This allows trunking efficiency and makes it more likely that a free channel will be available. In many cases, a single base station will not be able to cover the entire service area. If the uncovered area is limited to relatively small areas, such as in the shadow of a building, a remote radio port can be provided to illuminate this area. Since hand-held terminals usually have lower power than mobile terminals mounted in vehicles (due to battery and safety restrictions), mobiles can receive signals at greater ranges than hand-held. Portable vehiclemounted repeaters can therefore be used to provide hand-held coverage to users working near to their vehicles. This mode of operation is commonly used by the emergency services.

2.2 PMR Systems[1]

9

Fig. 2.4. Using a radio port to fill a coverage black spot

Fig. 2.5.

If larger areas have to be covered, several base stations must be used as shown in figure 2.6. If only a relatively low capacity is required, these can all transmit the same signal in a system known as simulating, and the system acts in the same way as one large cell. In analogue systems the frequencies used in the different cells vary by a few hertz which reduces problems in the overlap regions that receive signals from two or more cells. In digital systems, this is not possible, and systems have to be carefully designed to ensure that terminals can receive an adequate signal in the overlap region.

10


Fig. 2.6. Wide area coverage using several base sites

Larger capacity systems would require the use of multiple cellular reuse. Such systems are considerably more complex than other configurations, requiring switching between the base stations and handover of mobiles between cells as shown in figure 2.7. However, large PMR operators, and PAMR operators, need to use cellular configurations to give them sufficient capacity. Even large PMR or PAMR systems do not have as much traffic as public cellular systems, and so will have a relatively flat architecture compared to the complex hierarchical network architecture of GSM. The next section compares PMR and cellular operation more generally. A more detailed comparison will be given when TETRA will be compared technically with GSM in chapter 3.

Fig. 2.7. Cellular PMR Configuration

2.2 PMR Systems[1]

11

2.2.2 Comparison Between PMR and Cellular [2] PMR systems are in many ways similar to the public cellular systems. However, there are some significant differences between the two types of system which means that their design requirements are very different. The differences contribute to the security advantages of PMR over the cellular public systems as we shall see in chapter 3. The main differences between PMR requirements and the requirements of cellular systems are as follows. Group calls.

Cellular users have a much lower requirement for group calls than PMR users, and such requirements can usually be covered by having some sort of conferencing facility to link calls. PMR systems, on the other hand, must have flexible group call facilities, including allowing parties to enter and leave groups, and the ability to contact all users in a particular area. Dispatcher operation

Many PMR systems have a centralized dispatcher controlling and monitoring the system. This facility is not required in a cellular system. Decentralized operation

PMR systems are often required to work in a direct mode, where mobiles contact each other directly rather than via fixed base stations and network infrastructure. This allows operation outside the coverage of the fixed infrastructure and also in an emergency. Cellular systems must route all communications through the fixed infrastructure to allow for control and billing. Fast call set-up

Cellular users dial a number, and wait for their call to be connected. This may take tens of seconds depending on the call’s destination and call handling issues such as billing. In contrast, PMR users with a push-to-talk expect to do exactly that -press and talk- without delay. Supplementary services

Supplementary services are additional call services over and above the basic communication service. Examples include call forwarding for a voice call. PMR users are more likely to want supplementary services

12


tailored to their particular needs, such as variable priorities for different users, the ability to break into or monitor conversations, and so on. Cellular operators, on the other hand, have a much broader user community and will wish to offer a fixed range of simpler services which they can be confident will be commercially viable. Traffic patterns

With a PMR system which operates without dialing (i.e. a push-to-talk to contact the dispatcher or other users), calls are very short, consisting of a sentence or two. Usage regulations request a limit on shared PMR channels of 15-20 seconds, and the system may include a time out limiting the length of activity periods to 30 seconds or one minute so that one user is not able to hog a channel. In contrast, cellular calls will consist of a conversation, and so be longer. The average length of cellular calls is just under two minutes. Another difference between PMR and cellular is the destination of calls. Most cellular calls originate or terminate outside the mobile network, with only a small proportion of mobile to mobile calls within the operator’s network. On the other hand, PMR calls are usually intended for other users on that network, and the facility to route calls to other networks may even be absent as shown in figure 2.8.

Fig. 2.8. Sources and destinations of calls in PMR and cellular networks

2.2 PMR Systems[1]

13

Capacity

Cellular operators have a fixed allocation of frequency and they wish to maximize the number of users on the system to maximize their revenue. Their user base is large compared with their spectrum allocation and there is therefore an incentive to provide a large number of base stations, small cells, high re-use and efficient air-interface techniques to increase the number of simultaneous users supported. A PMR system is likely to have a much lower user base and the traffic is lower due to shorter calls, so capacity for the PMR operator is not likely to be an issue. PMR operators with their lower capacity requirements will want to minimize infrastructure costs, and so will have much larger cells, in the order of tens of kilometers. Cellular operators usually have cells limited to a few kilometers in radius at most. Capacity does affect PMR operators in another way when it comes to obtaining licenses. In many urban areas there are so many PMR operators that there are no spare frequencies and channels have to be shared. A PAMR system allows trunking of calls and results in a more efficient use of the spectrum. Frequency planning

In a cellular system, frequencies are planned throughout the whole system. This is not the case in PMR, where frequencies will be allocated to users for specific areas and there may be no co-ordination between users in a particular area. This means that a PMR system must obey strict interference limits with regard to neighboring earners, whereas cellular systems can tolerate adjacent carrier interference because neighboring cells can be planned with this in mind. Control, billing and authentication

In a cellular system the user is authenticated and billed for each call. This is in contrast to a PMR system where permitted users may use the system at will. The PMR operator has to pay for the infrastructure but this is effectively a standing charge and there is no per call charge. Relationship between the service provider and the user

In a PMR system the users are providing their own service, or will employ someone to provide the service on their behalf. The quality of the service is therefore directly within the control of the user. A cellular system

14


provides a standard service, though perhaps with some amendments. The user therefore has much less control. A PAMR system falls somewhere between these two extremes. Coverage

A cellular operator will provide coverage where it is economic to do so, which is where people who want to make mobile phone calls are. Cellular operators normally quote coverage in terms of the percentage of the population rather than the land mass, as complete coverage of the land mass of a country would be extremely expensive, and unless external factors such as government support are involved, may not be undertaken. On the other hand while cellular users may be able to operate their system at capacity in some areas, judging that the extra infrastructure costs would not be recovered by the additional traffic served. PMR operators may not have the option of dropping or queuing high priority calls, and will therefore have to provide additional capacity to meet worst-case, rather than average, traffic load. PMR operators usually require coverage over predefined areas of operation. Cellular users must be able to use their phones over as wide an area as possible, including internationally. PMR users may not be interested in use outside their specific location, although in the case of police services or truck drivers this might still be a considerable area requiring roaming between mobile networks. It is then obvious that there is need for the following peer standards.

2.2.3 PMR Standards [1] The Need for and Development of Standards With the move towards digital PMR systems, there has been a trend away from proprietary systems toward a public standard with which equipment must conform, thus allowing equipment from different manufacturers to be used together. Moves towards public standards have come from manufacturers and operators, as in the case of TETRA, or the user community, as in the case of APCO25. The attitudes of governments to the standardization process are quite varied. A hand off approach has been taken in the United States of America, where it has been decided not to insist on the APCO25 system but to allow the market to dictate which system is used. In contrast, in Europe, the European Commission is far more proactive in setting standards and even defining them at a technological level through

2.2 PMR Systems[1]

15

ETSI. This insistence that for certain government contracts systems conforming to ETSI proposals must be used is one of the reasons cited for the opening of the Matra PMR system resulting in the TETRAPOL standard. Public standards have a number of advantages, as was shown by the success of GSM in the mobile radio sector. Public standards enlarge the market, allowing an economy of scale as well as opening the market for niche players in more specialized areas. All the various standards include defined interface points allowing users to source different parts of the system from different suppliers. As well as forcing more competition between providers, it means that suppliers are no longer required to produce all the components of the system, although turnkey .solutions are still certain to be required by some users. Public standards also allow users more freedom to move equipment between networks. This is less of an advantage than it would be in the case of public cellular systems, where some users want a high degree of mobility and roaming between networks. However, it can still be seen to be an advantage to many PMR users, especially those, such as the emergency services, who co-ordinate with each other. Analogue PMR

Early PMR systems were analogue and proprietary. However, a wish to share infrastructure costs and a need to share spectrum led to the development of trunked radio systems, and with this development came the need for standards so that equipment could be sourced from different suppliers. The earliest such system, which is still available from a number of different suppliers, is LTR (logic trunked radio) developed by E F Johnson. A number of these systems are in operation worldwide. Another major analogue PMR standard is MPT1327 (Ministry of Post and Telecommunications), which developed in the UK in the late 1980s, but has been adopted by manufacturers and implemented worldwide. It is the most widely used PMR standard, common everywhere except the United States of America, where proprietary systems by Motorola, and to a lesser extent Ericsson, dominate. Although MPT1327 is more complex and expensive than some simpler analogue systems, it is relatively efficient in terms of spectrum use (digital systems are still better by a factor of two), and offers some data capabilities as well as the normal PMR voice call features such as group calls, fast call set-up, and priorities. PMR networks are expensive, and decisions to replace or upgrade are not taken lightly. Analogue systems arc likely to remain until capacity, maintenance or required features force a replacement. At the time of writing in 1999, manufacturers of LTR and MPT1327 equipment were still

16


promoting their analogue systems, in particular for use in countries or areas without spectrum capacity constraints. Digital PMR [4]

Digital systems offer a large number of advantages over analogue systems. A major advantage is the ability to recover the signal completely as long as the noise level is below particular threshold. This compares with the analogue case, where noise is always and degrades the quality of the signal. There is a disadvantage in that when level approaches the threshold of a digital system, the system performance falls off very rapidly, whereas in an analogue system the quality falls off steadily, giving clear of the system’s limits as shown in figure 2.9.

Fig. 2.9. Comparison of analogue and digital speech quality ith differing signal to noise level

Additional advantages relate to the sending of data, which can be sent directly in a digital system the requirement for a modem, and for trunking, as a digital signal can be manipulated more easily than an analogue one. While digital modulation is more easily than analogue systems, the transformation of a speech signal into digital form use of very efficient compression techniques so that the spectrum required for a speech signal is lower with digital modulation and good speech coder than with an modulation scheme.

2.2 PMR Systems[1]

17

Digital systems are more complex, and therefore more expensive. However, the increased flexibility, availability of services, and efficiency, in combination with the increased quality of service, means that the PMR market is now moving towards digital in the same way as the cellular market five years ago. A short description of the ones which made an impact are summarized in the following. 1. EDACS

EDACS (Enhanced Digital Access Communications System) is a proprietary digital PMR system from Ericsson. The first systems were installed in the late 1980s, and the system has found application in the military field, as well as its principle use in public safety. When the system was launched, a major selling point was its data services, which were unusual for a mobile radio system of that time, and the system has achieved considerable success, particularly in the USA. 2. Geotek-FHMA

Geotek-FHMA is a digital system which uses slow frequency hopping on an FDMA structure. The technology employed is novel for the civil mobile radio environment, being more common for secure military communications. As well as developing the system through its Israeli subsidiary, Geotek operated a limited number of digital networks itself in the USA, but the system suffered from limited take-up. A link up with IBM in 1997 failed to raise fortunes, and Geotek withdrew from digital network provision in 1998. The system itself, which has been installed in about half a dozen countries, is promoted as the PowerNet system for public safety applications along with Rafael, the Israeli defense firm which was a partner in its development. National Band 3, a UK PAMR operator owned by Geotek, was going to adopt Geotek-FHMA, but this network is now owned by Telesystem International Wireless of Canada, which through its Dolphin Telecom subsidiary is using TETRA for new digital PAMR operations in the UK and France 3. APCO25

APCO25 was an initiative by the Association of Public-safety Communications Officials - International, Inc. (APCO) to try to create a standard PMR system for public safety applications. While the emphasis in Europe has been to create a cross-border standard for such systems, the United States has a more market-orientated culture and different countries

18


may have incompatible systems from different suppliers. The idea of a federally required standard has since been watered down, but the system has continued to be developed with Motorola, which owns lights to much of the key technology, licensing this to other manufacturers. Motorola’s Astra system was the first APCO25 system, launched in 1996. Recently APCO25 and TETRA agreed to cooperate on future developments. In its current form, APCO25 is an FDMA system with 12.5 kHz carrier spacing, which is compatible with existing analogue channel spacing. However, future plans foresee halving the bandwidth requirements for speech channels to make more efficient use of spectrum. A narrow-band FDMA approach with 6.25 kHz carrier spacing has been proposed to allow this, but this is being reconsidered in the light of TETRA developments which may see a TDMA approach being employed for this development has had a presence in the market for a couple of years before the roll out of TETRA systems, and has been adopted by users in 15 countries, mainly in the public safety area. 4. TETRAPOL

Tetrapol faces a problem in terms of take-up in the EU, since although it is recognized by the ITU, it is not a formal standard approved by ETSI, and moves to convert the TETRAPOL PAS (Publicly Available Specification) to an ETSI standard have recently stopped. Most European Union countries are planning to use TETRA, although TETRAPOL has been recognized by Schengen Group along with TETRA, and it is in use by security forces in France, Spain and Austria, as well as other European countries such as Romania, Slovakia and the Czech Republic. 5. TETRA

The main focus of this book is the TETRA standard as used in security applications. More technical details are given in chapter 3 in the forum of a comparison between TETRA and GSM. the appendix we summarizing the technical characteristics of TETRA. TETRA was developed from the start as an open harmonized digital PMR standard within ETSI. As an ETSI approved system, it has a significant commercial advantage within Europe, both from the point of view of manufacturer and operator support, and from the point of view of governments, which in Europe will specify ETSI approved systems for their contracts. The wide user and producer base should provide significant economies of scale. However, the PMR market has a number of existing 2nt generation digital systems in operation already, and with 3ld generation cellular systems only a few years away it

2.2 PMR Systems[1]

19

is important that TETRA gets off to a good start if it is to establish a dominant position in the marketplace. TETRA is a feature-rich system, providing everything from specialized safety services to cellular operating modes. It also has a wide selection of data services. The trunked TDMA access technique allows more efficient use of the radio spectrum, but means that an operator must be assigned a minimum of at least four voice channels. However, the system has a number of operating modes, which allow wide area coverage with a single radio carrier without resorting to cellular frequency reuse schemes that would increase radio carrier demands still further. Of more potential concern is that the complexity of the system will make the infrastructure and terminals relatively expensive, which should be offset by the economies of scale if the system becomes popular. 6. Mobile Satellites

Although it is uncertain as to when satellite communications will be practical and economical for use by public safety agencies, it is critical to discuss these important emerging technologies in this handbook. The United States has a fleet of geosynchronous earth orbit (GEO) satellites at approximately 22,500 miles above the equator providing wideband transponders to connect telephone and television circuits around the world. There are several GEO systems used for general mobile services available today. However, they require a briefcase full of equipment, including a highly directional antenna. In addition, there is a delay of about 1/4 second for the transmission, which slows down interactive voice and data transmissions considerably. Because of this, the service is not yet appropriate for the use of simple handsets as used for cellular or PCS radio. 7. Voice Communications Satellites

Besides GEOs, medium earth orbit (MEO) and low earth orbit (LEO) satellites have been proposed for relaying radio transmissions. MEO and LEO satellites require less output power from phones and have less time delay than GEO systems. The relationship of GEO, MEO, and LEO satellites so that in a full implementation of GEOs, MEOs and LEOs, there will be three discreet bands over the eath where the satellites will be located. Iridium®[5].

In 1987, Motorola engineers proposed their Iridium satellite system for wireless communications to allow a person with a small handset anywhere on the earth’s surface to communicate with another person’s handset

20


anywhere else on the earth’s surface. This satellite system was the first of a number of systems that would not only receive signals from the earth (which are converted in frequency, amplified, and re-transmitted as commonly done in transponders) but would also contain switching and routing processors. Iridium system was constructed and functioned as planned; however, Iridium, LLC, filed for bankruptcy in February, 2000, because of a failure in their business plan. In December, 2000, Iridium Satellite, LLC was formed and acquired the operating assets of Iridium LLC including the satellite constellation, the terrestrial network, Iridium real property and the intellectual capital. A new management team was installed and the company sold their services in March, 2002, to the U.S. Department of Defense as a stable customer. The Iridium system consists of 66 satellites placed in LEO orbits with seven spares to fill in should the company loose the service of a satellite. The system is composed of 6 planes of 11 satellites equally spaced in a low-elevation orbit with an orbit altitude of 421.5 nautical miles, as shown in figure 2.11. Each satellite provides a set of 48 separately controlled spot beams to cover the earth’s surface so that (with the 66 satellites) there will be 3,168 cells covering the entire earth. The system may be thought of as a type of cellular radio system where the “cellular base stations” and cells are constantly rotating so the earth signals are handed off from one satellite to another as they pass over an individual’s handset. L-band frequencies (1616 - 1626.5 MHz) are to be used for the communications between the earth and the satellites and the Ka-band frequencies (23.18 - 23.38 GHz) are used for intercommunications between the satellites. Ground segment frequencies to gateways and control facilities use Ka-band frequencies (downlinks, 19.4 - 19.6 GHz, uplinks, 29.1 - 29.3 GHz). Figure 2.10 shows Motorola’s concept of this system. Iridium will support voice and data up to 4800 bps.

2.2 PMR Systems[1]

21

Fig. 2.10. Iridium System Overview

Other voice satellite systems.

The other commercial LEO system in orbit is Globalstar, which is now a wholly owned subsidiary of Vodafone Group PLC. The system consists of 48 satellites allowing for seamless coverage anywhere on the earth. The system utilizes CDMA technology with path diversity and the company provides light weight, 12 oz. phones for voice communications. Other companies have stated an interest in LEO and MEO narrow band systems. Mobile Communications Holdings’ Ellipso™ and ICO Global Communications’ ICO satellites are in MEOs, spaced at about 6,000 to 10,000 miles above the earth’s surface. There are tradeoffs between the LEOs and MEOs. Far fewer satellites are required in the MEO system than in the LEO system, but higher effective power is required for transmissions by the subscriber units, and time delays are greater.

22


8. Data Communications Satellites

Since 1992, American Mobile Satellite Corporation, now Motient Corporation, has offered satellite service employing geosynchronous satellites. Motient has recently transferred it’s operating interest in satellite communications to a partnership called Mobile Satellite Ventures LLC (MSV). Motient retains approximately 25% interest in the partnership. The MVS system provides coverage to North and Central America, parts of South America and the Caribbean via a single geosynchronous satellite using “L-band” technology. Both voice and data may be handled on the same system, with communication of data up to 4800 bps. The equipment used includes both mobile and transportable units. The mobile units use a steerable antenna to allow use on a moving vehicle. The system provides three different services. The first is a satellite telephone service that allows calls to be made to any phone through the PSTN and unit-to-unit calls to be made through the satellite without use of any ground stations. The second service provided is a radio-like service that allows unit-tounit calls via a talk group. Satellite units can have multiple talkgroups, and operate using the system as a satellite-based trunked radio system. Some rural fire and EMS agencies use this system for radio communications over very large areas. In addition, several local and US government agencies use these talk groups to coordinate task force disaster responses. The final service is a packet data service. This service is relatively low speed and useful primarily for fleet tracking and equipment control. Motient provides dual mode services allowing mobile units to use their terrestrial service when within their coverage area and to automatically switch over to the MVS satellite system when the terrestrial system is not available. Wideband, data-oriented LEO and MEO PCS satellites are being studied and proposed at this time, as shown in table 2.2. These satellite systems will have the ability to carry high-speed data around the world at up to 10 Gbps.

9. High Altitude Long Endurance (HALE) Platforms and High Altitude Platforms (HAPS)

In this proposed network, relay of signals would be accomplished using large blimp-like repeaters at several miles (20,000 meters) above the earth. The devices would cost less than the big satellite systems and could be recalled to earth for maintenance. Multi-beam, phased array antennas

2.2 PMR Systems[1]

23

would support both mobile two-way communications and broadband video. Although not considered HALE/HAPS, the U.S. is presently performing surveillance over the U.S.-Mexican boarder using low altitude tethered balloons carrying electronic equipment. Four types of HALE platforms have been proposed, which include helium-filled robotically controlled dirigibles stabilized by ion engines units powered by solar or fuel cells; piston-driven platforms; and jet enginedriven platforms. The biggest challenge faced by all of them will be power requirements versus refueling requirements. The first two types need little or no refueling but may not produce the transmit power needed, whereas the latter two types will have plenty of power but will need to be refueled every few days. Sky Station International was the commercial initiator of this technology in the United States and filed with the FCC in March 1996 for use of the 47 GHz band. Sky Station claims a blimp repeater can offer many advantages over satellites, including less time delay and lower power at a considerably lower cost. The concept was also introduced at the 1997 World Administrative Radio Conference, and a portion of the 47 GHz band was tentatively allocated. The 47 GHz band is severely limited by rain, so space diversity ground circuits will most likely be required. The basic concept is to have “very high antenna towers” allowing for very wide-area communications. This might be an alternative to backbone microwave terrestrial systems. Sky Station indicated that one could start with communications in local areas, expand to regional areas, and eventually cover the country. The FCC has made no decisions at this time. The concept has many technical and political challenges, and its development should be interesting to watch as it evolves. Since the first edition of this book much greater consideration has been given to HALES/HAPS by many countries throughout the world. NASA has proposed a schedule for testing systems by 2003 using manned and unmanned aircraft and balloon type platforms23. The HALE/HAP systems at a height of 25 Km appears to have the least amount of wind speed and a coverage is about 200 Km. Among the multitude of technical problems to be solved are: 1. Developing stability systems to hold the HALE/HAPS at station keeping locations and stabilize microwave antenna positions. 2. The testing of aerodynamics and aircraft structures. 3. The development of additional light weight, high strength materials. 4. Making sure the altitudes of HALE/HAPS will not interfere in any way with normal air or military air travel.

24


10. Ultra Wide Band (UWB) Devices

Microcircuit advances in the last year or so have made it possible to create ultra wideband (UWB) radio and radar equipment having very narrow digital pulses, in the nanosecond range, to transmit and receive very high rate data information. The bandwidths are very large and cover a great amount of the licensed frequency spectrum. The FCC and NTIA have studied and made tests to determine that the use of UWB at low power levels will not cause objectionable interference to those licensed services. In February, 2002, the FCC enacted rules under Part 15 to assign certain frequency ranges for UWB and to quickly allow for the development of commercial devices using this new technology. The UWB research has already yielded a number of new devices which will assist public safety groups as soon as the equipment is developed. These include: •

•

•

•

•

High Speed in Building Radio Communications - High speed digital transmission with rates in the gigabit range for computer networks using work stations or handheld devices within buildings. The transmissions must take place in the 3.1 - 10.6 GHz spectrum. Building Penetration Radar - Radar has been developed for firefighters to look into buildings through walls to find the position of people trapped during a fire. Similarly police surveillance may utilize this radar to determine the number and locations of people within buildings. Operation is limited to law enforcement and fire and rescue operation. The radar emissions must be kept within the 3.1 - 10.6 GHz frequency domain. Ground Penetrating Radar Systems - Public safety personnel may use ground penetrating radar (GPR) to determine the location of buried objects including the locations of people within the rubble of fallen buildings. Operation of the GPR is restricted to law enforcement, fire and rescue operations, scientific research institutions, commercial mining companies and construction companies. GPRs must be operated below 960 MHz or between 3.1 - 10.6 GHz. Surveillance Operations - Surveillance operations, as opposed to the wall penetration systems, are defined by the FCC operate as “security fences” to establish stationary RF perimeter fields to detect the intrusion of people or objects. Operations of these devices are limited to law enforcement, fire and rescue organizations, public utilities and industrial entities. The frequency band established is 1.99 - to 10.6 GHz. Vehicular Radar Systems - Licensed in the 24 GHz band, this UWB technology using directional antennas on road vehicles will detect and locate

2.2 PMR Systems[1]

25

the movement of objects near the vehicle to enhance crash avoidance systems, improve airbag activation and suspension systems that will respond better to road conditions. All the above uses of UWB are licensed under Part 15 of the FCC Rules and are subject to power restrictions as well as the frequency restriction discussed previously. 11. Software Defined Radio (SDR), Cognitive Radio, Wireless Mesh Networks

Public safety radio systems have changed over many years from simplex radios, to single repeaters, to trunked radio systems in both analog and digital configurations. Current radio systems make use of digital circuitry to emulate a number of these earlier configurations, allowing for efficient interoperability within single frequency bands. However, the next big change in radio design is the use of software to dynamically change a radio’s configuration to emulate a multitude of protocols and modulation waveforms using the same hardware. Microprocessors are already used in today’s radio systems at specific frequency bands to set up the transmitting and receiving frequencies, as well as for other functions including user configurations for different talk groups. Software Defined Radio (SDR) promises to integrate entire radio functions (including transmitting, receiving, signal processing and networking) to allow for specific hardware to be dynamically reconfigured to all types of public safety radio systems across multiple bands with a simple change of the “channel” switch. The resulting product will be the ultimate solution to the interoperability problem, giving the field officer a radio “on-the-belt” that can access many different systems on different bands, depending upon the configuration authorized by his agency. Additionally, this radio hardware platform should be easily upgradable to new technologies as they develop, reducing equipment obsolescence as new features, functions and systems are introduced. An example of early software controlled radio is the use of multimode cell phones which allows a subscriber to automatically switch from the 800 to 2000 MHz frequency bands and emulate the present TDMA, FDMA, and CDMA standards without using several different cell phones. The third generation of cell phones is already utilizing as many as seven independent standards to automatically accommodate different transmission modes. Although the software for these technologies is embedded in chips, SDR promises to allow dynamically updated changes so that it will not be necessary to purchase new hardware every time an update to more efficient technology is made.

26


These systems can support a number of military and public safety waveforms, and covers the public safety bands from 30 MHz to 512 MHz. While the radio does not yet protocols, there is significant interest by the manufacturer and governmental agencies to add these protocols to the radio. While this radio currently costs about twice as much as a similarly featured single band radio, if it replaces radios on three different bands (as it is capable of doing), it today provides significant savings. As with all new technologies, prices should drop significantly as market penetration increases. There is a trend, in general, that these systems as they are also coined as cognitive radio, to be combined with PMR and have as a building block the TETRA system. This bring us the prospect of using the rich functionality of TETRA and create TETRA extensions to include integration with WLANs and wireless mesh networks as explained in chapters 6 and 7 be able to use TETRA as the integration component of any possible technology that is being used for emergency, public safety and security communications used even in antiterrorist applications. Therefore the title of the book: TETRA – A Global Security Tool is well justified. 12. Voice Over Internet Protocol (VoIP)

Using the Internet for wireless information applications is one of the latest technology developments to hit the telecommunication world. Voice delivered using the Internet Protocol, or IP, is simply a way of sending information from one device (a desktop computer, for example) to another (radio) over the Internet. To do this, voice information is converted into digital form and then sent in discrete packets over the internet to a receiving device on the other end Changes in technology enable more information to be sent at higher speeds, including voice, fax, video and data through a single large pipeline. With the passage of IEEE’s 802.11e standard, more network managers will be administering wireless voice over IP. This can mean private radio or cellular wireless or both. The standard is focused on supporting video on demand and audio on demand. Multimode devices (such as NIC cards) are being developed that will work with a choice of 802.11e wireless LAN or cellular wireless LAN. Other researchers are working on multimode for 802.11 and CDPD. These multimode devices will likely be targeted towards users, such as business people in airports, who need to make cellular voice calls as well as send data over the Internet using a wireless IP link.

2.2 PMR Systems[1]

27

In public safety applications, portable radios could receive pager-like text messages, reducing the demand on voice traffic, mug shots could be sent from headquarters to the field officer, video footage can be sent from a crime scene to a dispatch center for assistance in resolving highly volatile situations, and GPS tracking is available for the added safety of officers in the field. 13. Private Wireless Wideband Data Systems

Private Wireless Systems can be extended to cover mobile video, voice and data transmissions simultaneous at as low as 460 kbps data takes. One such application has been demonstrated successfully by Motorola in Florida by the “Greenhouse Project” This experiment proved that applications conducted at one’s personal computer may be accessed wirelessly in the field and teleconferencing may be conducted wirelessly from a field facility to another field or fixed facility. Tests were being conducted from patrol cruisers, surveillance vans, ambulances, fire engines and fire district vehicles equipped with Greenhouse equipment. Greenhouse supports the following technologies and applications: • • • • • • • • • • • • •

Video (Streaming IP video: 2-way video, 1-way video, video pull, video push) Voice (Voice over IP - Internet Protocol, Full Duplex - both users can talk at the same time) Data (high-speed mobile access to intranet and internet) Some applications include: Automatic Vehicle Location (AVL) through GPS - vehicles locations appear on map Electronic Mail - instant messages including attachments Computer Aided Dispatch - facilitates quick deployment of public safety officials National and State Crime Database Access - ability to check drivers licenses, etc. The ability to distribute a picture of a missing child, or criminal suspect/sketch to all equipped vehicles in the field Robbery videotapes can be distributed shortly after an event Enable fire department access to building plans and hydrants Transmit fingerprints Transmit live video feeds for police officer pursuits Remote situation analysis

28


For example, at a crime scene, an officer may take a digital camera mug shot and crime scene pictures; digitize finger prints; select driver’s license signature and picture information and send messages (wirelessly) over such a system to obtain crime analysis data from NCIC as well as infrastructure information if required.

2.3 PMR Limitations [4] Conventional PMR communications systems as we have seen bring many benefits to user organizations but they also create many operational problems for the radio user. Even the most simple of systems comprising a single base station operating on one RF channel and using open channel ‘all informed net’ operation has some annoying operational problems/limitations. When the size of a conventional system increases and utilizes several base station sites with more than one base station per site, the annoying operational problems are multiplied. Moreover, when large scale PMR systems of different technologies and features are use in security applications and especially when infer organizational or infer country applications are involved interoperability is a major issue and we shall see in section 2.3.3. For example, operational problems reported are: • • • • • • •

Edge of coverage voice quality Contention Manual switching of channels Inefficient channel utilization Lack of privacy Abuse by radio users Limited data throughput

A detailed description covering each of these operational problems is provided in the following text. 2.3.1 Edge of Coverage Voice Quality Analogue wireless systems were designed for noise limited range performance, as a trade-off between communication quality and infrastructure cost. This cost versus communication quality trade-off was not unusual as organizations traditionally procured their own private networks. For example, police forces, fire brigades, ambulance services, municipalities, gas

2.3 PMR Limitations [4]

29

utilities, electricity utilities, water utilities and public transportation organizations often had their own private PMR networks covering similar geographic areas, with many of the base station sites being shared to provide coverage. This cost versus communications quality meant that voice quality at the edge of RF coverage areas was often poor, as received signal levels decreased and noise levels increased. The resulting poor voice quality meant that messages were often repeated and in some cases lost. Also, this poor voice quality placed radio users under increased stress in ‘listening out’ for voice messages in noisy background signals. Contention

Radio users wanting to access the system needed to wait until the channel was free of traffic. This meant that they had to monitor the channel waiting for an appropriate time to initiate a call. During busy periods many users were waiting to access the network resulting in several users transmitting at the same time when the channel became free. This simultaneous transmission often corrupted received messages and resulted in users having to contend with each other again in order to access the system. As a consequence, many lower priority calls were never established.

Fig. 2.11. Contention between users accessing a Base Station

30


The diagram in Figure 2.11 shows contention clashes between three user groups, A, B and C attempting to access a single Base Station (BS) operating on frequency Fl. The high probability of contention clashes during busy periods led to user frustration and also limited the number of radio users that could be supported, which were often as low as 15 to 20 per communication channel. Manual switching of channels

In systems with more than one base station site to provide wide area coverage, radio users needed to switch to a different radio channel when moving from the RF coverage area of one base station site to that of another. This manual switching of channels was relatively simple to do if the radio user knew when they were out of RF coverage and also knew which channel to switch over to for service. But this need placed the burden on radio users to monitor the channel for lack of coverage and then make the channel changeover decision. As a consequence, radio users were often out of communication without realizing it, until they tried to initiate a call. Inefficient Channel Utilization

When there were many radio users on a system, more than one base station was provided at a base station site to provide the required capacity. However, for ease of operation and because of technology limitations, radio users in different disciplines of an organization were given only one channel to use, even though more than one channel was available at the base station site. The diagram in Figure 2.12 shows three base stations, each supporting three independent user groups -groups A, B and C operating on BS Fl, groups D, E and F operating on BS F2 and groups G, H and I operating on BS F3. From the diagram it can be seen that user groups A, B and C attempting to access BS Fl are experiencing contention clashes and resulting poor grade of service whereas Group D has gained access on BS F2 and is experiencing a good grade of service. Base Station F3 is not being utilized because no user groups need communication services at that moment in time.


31

Figure 2.12. Representation of a conventional multiple channel PMR system

This would mean that when one channel was busy with several users waiting to gain access, other channels would normally be free of traffic and could have been used to lessen the load on the busy-radio channel. As a consequence, radio users often encountered unnecessary contention and wasted valuable time through inefficient channel utilization. Lack of Privacy

Because every radio user listening on a radio channel could hear what everyone else was saying, as anyone can with an easily obtained radio receiver, communications privacy was virtually impossible. This situation had particular implications for the emergency services as well as some commercial organizations. Abuse by Radio Users

Even in the most disciplined of organizations there will always be individuals who will abuse the system if they can get away with it. This is unfortunately true for radio users. Common examples of abuse by radio users are:

32

• • •


Ignoring messages and saying that they must have been out of coverage Saying they were in one location when they were in another Giving verbal abuse over the radio channel knowing that they could not be identified

Limited Data Throughput

The traditional method used by regulators to increase RF spectrum efficiency and utilization was frequency division. For example, over time 50 kHz channels were replaced with 25 kHz channels, 25 kHz channels replaced with 12.5 kHz channels and in some cases 12.5 kHz channels replaced with 6.25 kHz channels. As far as occupied channel bandwidth was concerned, dividing channels by two effectively doubled the number of channels available for a given amount of RF spectrum. Reducing channel bandwidth for voice communications had little effect on quality. However, as a consequence of using narrow band channels, mobile data throughput was limited to relatively low-speeds. The laws of physics meant that the narrower the channel bandwidth, the lower the data throughput. From a user’s operational aspect, this limitation in data throughput meant that practical applications were limited to status messages and short data messages. Users were also particularly concerned about mixing voice and data messages on the same channel, as the data messages were audible and annoyed users. Problems Solved

Fortunately, identifying problems is a precursor to solutions being provided. In the problem areas described previously there are solutions that have evolved to overcome some of these problems. For example, Continuously Tone Coded Sub-audible Squelch (CTCSS), also known as Private Line (PL), helped to provide some degree of privacy between radio user organizations operating on the same system, but not against eavesdroppers with radio receivers. Press To Talk (PTT) inhibit, automatically enabled when the channel is busy, helped to minimize transmission clashes and reduce the number of repeated messages. Selective signaling, such as 5/6 tones sequential, reduced verbal abuse because users can be automatically identified at the start of each transmission. Also, selective signaling provided some degree of individual user privacy, but again not against eavesdroppers with radio receivers.


33

Automatic Vehicle Location (AVL) provided by Global Positioning System (GPS) and other location technologies, has greatly assisted the operational efficiency of many organizations and has also prevented false location information being provided by radio users. By far the most beneficial technology solutions are those of trunking and digital wireless communications. Because of the importance and significance of trunking and digital cellular systems in solving many conventional PMR operational problems, a comparison is made between PMR and cellular as well as TETRA and cellular in separate chapters. The clear advantage of TETRA then is proven in chapters 5,6,7. 2.3.2 Requirements of PMR Services The requirements of a private mobile radio system can be summed up very simply as giving the ability for users to communicate with each other reliably. More specifically, it is possible to identify a number of key requirements of PMR users. In no particular order, these are: Reliability

Many PMR services are used in safety critical systems. One advantage to the user of being involved in the operation of the service is that they are in the position to ensure reliability and are not dependent on other operators. The lack of public cellular systems to guarantee quality of service or grade of service in all circumstances, or their unwillingness to take liability for safety critical services, may force the use of a PMR system. A survey that looked at the importance of PMR features found that service availability was classed as “extremely important” by two-thirds of those questioned, the highest proportion of any requirement. Speech and data transmission capability

Mobile data services are increasingly being used for tracking, telemetry or information updating services. Examples of innovative data service use include BT, a national telecommunication operator, which sends daily work orders direct to repair technicians so that the working day can start at the first job rather than with a trip to the depot. Simoco is conducting trials with Langdale Ambleside Mountain Rescue Team in the UK on transmitting medical telemetry, including still images, video, text messages and GPS data to assist in rescue operations. As data services develop, so will

34


the applications which make use of them. A flexible data service provision is therefore essential. In the survey [6], almost 80% of users classed data communications as important, with over half of these saying that it was “very important”. 10% of respondents classed data calls as “not important”, but all respondents classed speech calls as important to some degree. Centralized and decentralized operation

In many businesses, PMR is used to organize users, and a central dispatch point is therefore required. However, it may also be important that users are able to contact with each other in the absence of a central control point or even any infrastructure at all. Again the survey [6] found almost 80% of users classed direct mode operation as important, with over half of these saying that it was “very important”. Point-to-point, group calls and broadcast calls

If PMR systems are used, a flexible group call structure is essential so that users can share information directly rather than having to relay it via others. Therefore, group calls, calls involving a number of defined users, and broadcast calls, where the call includes all terminals, are required in addition to point-to-point (single terminal to single terminal) calls. Fast call set-up

Rather than dialing a number to set up a call, with the called party answering a phone, PMR systems usually have a “push-to-talk” to activate a call to the dispatcher or user group, with the receiving terminal annunciating the message without an answering procedure. Calls may therefore consist of a sentence or two, and users expect to be connected to the called terminal without delay. This is particularly important in the emergency services where the radio may be used to give urgent commands and the dropping of the first few words of the message due to delaying in setting up the call might have serious consequences. Good coverage

Professional mobile radio users usually have less choice as to where to make a call than a cellular user. The call location is often stipulated by the location of the work the user is undertaking. In the case of a utility this may mean having good coverage over a wide area, and for public safety


35

users, constraints can be even more severe. For example, a mountain rescue service may require coverage in areas where public cellular systems would not be provided, but even in more benign radio environments, as well as overall area coverage, the absence of black spots within a covered area is also very important. Long battery life

User maintenance costs money in terms of lost work time in PMR systems, and reliability of service is also important. This compares with public cellular systems where the users are responsible for battery charging. Flexibility

Flexibility takes many forms. Flexibility with regard to services has already been covered, but another input aspect of flexibility is the ability of the system to change with the developing needs of the operator. In particular, the system should be scalable so that growth can be handled, and sufficiently adaptable to allow new services, which were not anticipated when the system was installed, to be added later. Businesses will not want to invest large sums of money in a system which cannot be modified easily once installed. Low total cost of ownership

Companies using PMR systems will consider cost over the entire life of the equipment, including capital costs for the infrastructure and maintenance costs for the equipment in addition to the “headline” cost of the terminals themselves. Unsurprisingly, no respondents to the survey in (61 classified costs as unimportant, with 95$ classifying them as “important” or “extremely important”. A number of other requirements may not be necessary in all cases but will be needed by a large number of users. Any PMR system will therefore have to take them into account. Security

Many PMR users have a requirement for high levels of security. Security takes a number of different forms, both in terms of reliability of operation and protection of the transmitted information from tampering and interception.

36


Call priorities

PMR operators may wish to be able to differentiate between users to give different call priorities or qualities of service to different user or call types. For example, an emergent call may be able to pre-empt other call types to gain access to the network. Communication between networks

Many companies operate over large areas or with several sites. They may not want to provide the complete network themselves or they may use different networks on different sites due to equipment replacement cycles or regulatory restrictions. Their PMR networks may therefore have to communicate with each other. Also, in many circumstances, communication with general telephone or data networks is a useful Infrastructure Ease of licensing

This issue involves not just the bureaucratic process of obtaining permission to use a radio channel but also the issues of the availability of channels and any co-ordination which may be required with other users in the same area. The problem of licensing hundreds of different users operating in numerous different areas is much more complex than that of organizing a small number of national cellular operators. It is only possible if the PMR radio channels are as self-contained as possible from the point of view of interference with other users. In-house control of system

Almost two thirds of PMR users preferred to control their own network, with less than one sixth thinking that control was not an issue. An obvious reason for this is to ensure security, but other advantages relate to cost control and service guarantees. One requirement not included above is that of capacity, or the efficient use of radio resource. In tact, capacity is not normally an issue to PMR users due to the length of the call, and since licenses have been relatively cheap in most countries. PMR users have only been concerned at a more general level as channels become scare and have to be reused more frequently in high traffic areas such as cities. Of more general concern is the fact that the division of the available spectrum between different PMR users means there is little trunking efficiency. This results in there being fewer channels than required in most large urban areas. The regulatory


37

authorities are therefore likely in insisting that PMR systems are spectrally efficient and use a narrow carrier spacing. 2.3.3 Interoperability [6]

Introduction

As PMR technologies become more far-reaching and interconnected, interoperability has become critical. Interoperability to achieve information superiority is the keystone on which future security systems, logistic, and other government systems will be based on. Interoperability is, therefore, the foundation of effective joint, multinational, and interagency operations. We have seen that for an effective, overall design of a system that will satisfy modern security requirements must be able to utilize the resources of all subsystems that have been used so for emergency security or public safety implementations. In order to integrate all of them for a particular and multifunctional application we can use show the TETRA can accomplish this objective and serve as an integrating system as will explain in later chapters. The general framework that covers such operation is coined interoperability which will be described in the section in more general terms. Currently, there is a tendency to concentrate on the mechanisms that various systems use to interoperate. However, focusing solely on mechanisms misses a larger problem. Creating and maintaining interoperable systems of systems requires interoperation not only at the mechanistic level, but also at the levels of system construction and program management. Improved interoperation will not happen by accident and will require changes at many levels. While many systems produced by security agencies can, in fact, interoperate with varying degrees of success, the manner in which this interoperation is achieved is piecemeal. In the worst case, interoperability is achieved by manually entering data produced by one system into another—a time consuming and error-prone process. cross-organizational home. Although technical interoperability is essential, it is not sufficient to ensure effective operations. There must be a suitable focus on procedural and organizational elements, and decision makers at all levels must understand each other’s capabilities and constraints. Training and education, experience and exercises, cooperative planning, and skilled liaison at all levels of the joint force will not only overcome the barriers of organizational

38


culture and differing priorities, but will teach members of the joint team to appreciate the full range of Service capabilities available to them. A proper definition of interoperability is a prerequisite to any discussion followed. Defining Interoperability

There is a need for precise definition of interoperability, because the term can have various interpretations in different contexts. For example, interoperability between a field commander’s planning systems and a weather system may be addressed via a simple broadcast email. In contrast, radar reports of objects in the environment that must be shared between complex systems like AWACS may require frequent, automated updates of complex information. Some of the difficulty associated with defining interoperability is reflected in the many definitions that exist. For example, the IEEE has four definitions of interoperability: • the ability of two or more systems or elements to exchange information and to use the information that has been exchanged. • the capability for units of equipment to work together to do useful functions. • the capability, promoted but not guaranteed by joint conformance with a given set of standards, that enables heterogeneous equipment, generally built by various vendors, to work together in a network environment. • the ability of two or more systems or components to exchange information in a heterogeneous network and use that information. Security agencies also use multiple definitions of interoperability, several of which incorporate IEEE definitions: The ability of systems, units, or forces to provide services to and accept services from other systems, units, or forces, and to use the services so exchanged to enable them to operate effectively together. The condition achieved among communications-electronics systems or items of communications-electronics systems equipment when information or services can be exchanged directly and satisfactorily between them and/or their users. The degree of interoperability should be defined when referring to specific cases. For the purposes of this instruction, the degree of interoperability will be determined by the accomplishment of the proposed Information Exchange Requirement fields. The result that is strived for is (a) Ability of information systems to communicate with each other and exchange information. (b) Conditions, achieved in varying levels,


39

when information systems and/or their components can exchange information directly and satisfactorily among them. (c) The ability to operate software and exchange information in a heterogeneous network (i.e., one large network made up of several different local area networks). (d) Systems or programs capable of exchanging information and operating together effectively. We may never have agreement on a precise definition due to differing expectations that are constantly changing. New capabilities and functions continue to offer new opportunities for interactions between systems. For the purposes of this report, we define interoperability as: The ability of a set of communicating entities to (1) exchange specified state data and (2) operate on that state data according to specified, agreed-upon, operational semantics. Models of interoperability

There exist a number of models of Interoperability which will be analyzed below. For the purposes of this book, we shall stress the technical model. Levels of Information System Interoperability

A widely recognized model for system of systems interoperability is Levels of Information System Interoperability (LISI). LISI focuses on the increasing levels of sophistication of system of systems interoperability. Five levels are defined: Level 0 – Isolated interoperability in a manual environment between stand-alone systems: Interoperability at this level consists of the manual extraction and integration of data from multiple systems. This is sometimes called “sneaker-net.” Level 1 – Connected interoperability in a peer-to-peer environment: This relies on electronic links with some form of simple electronic exchange of data. Simple, homogeneous data types, such as voice, text email, and graphics (e.g., Graphic Interface Format files) are shared. There is little capacity to fuse information. Level 2 – Functional interoperability in a distributed environment: Systems reside on local area networks that allow data to be passed from system to system. This level provides for increasingly complex media exchanges. Logical data models are shared across systems. Data is generally heterogeneous-containing information from many simple formats fused together (e.g., images with annotations).

40


Level 3 – Domain based interoperability in an integrated environment. Systems are connected via wide area networks. Information is exchanged between independent applications using shared domain-based data models. This level enables common business rules and processes as well as direct database-to-database interactions. It also supports group collaboration on fused information. Level 4 – Enterprise-based interoperability in a universal environment: Systems are capable of using a global information space across multiple domains. Multiple users can access complex data simultaneously. Data and applications are fully shared and distributed. Advanced forms of collaboration are possible. Data has a common interpretation regardless of format. Within a level, LISI identifies additional factors that influence the ability of systems to interoperate. These factors comprise four attributes: Procedures, Applications, Infrastructure, and Data (PAID). PAID provides a method for defining the set of characteristics required for exchanging information and services at each level. It defines a process that leads to interoperability profiles and other products. Scenarios depict the possible uses of LISI in different circumstances throughout the system life cycle. LISI focuses on technical interoperability and the complexity of interoperations between systems. The model does not address the environmental and organizational issues that contribute to the construction and maintenance of interoperable systems (e.g., shared processes for defining interoperability requirements and maintaining interoperability across versions). 2) Organizational Interoperability Maturity Model

The Organizational Interoperability Maturity Model (OIM), which extends the LISI model into the more abstract layers of command and control support, describes the ability to interoperate. It can cover cases where the organization depends, on one extreme side, on independent sub- groups and on the other extreme side on unified sub-entities. 3) Technical Architecture Reference Model for Interoperability

If we try to develop Technical Architecture Model for Interoperability, we should be able to define four degrees of Interoperability regarding data exchanges.


41

Degree 1 - Unstructured Data Exchange: exchange of humaninterpretable unstructured data such as the text found in operational estimates, analyses and papers. Degree 2 - Structured Data Exchange: exchange of human-interpretable structured data intended for manual and/or automated handling, but requires manual compilation, receipt and/or message dispatch. Degree 3 - Seamless Sharing of Data: automated sharing of data amongst systems based on a common exchange model. Degree 4 - Seamless Sharing of Information: universal interpretation of information through data processing based on cooperating applications. The degrees were intended to categorize how operational effectiveness could be enhanced by structuring and automating the exchange and interpretation of data. It is obvious that the organizational and technical model must merge as we shall see in the following model which is of major importance in the design of an integrated system that includes TETRA as a building block. 4) The System of Systems Interoperability (SOSI) Model

The models previously discussed address a range of interoperability issues from technical to organizational. The SOSI model, addresses technical interoperability and operational interoperability. However, SOSI goes a step further to address programmatic concerns between organizations building and maintaining interoperable systems. Interoperation among systems is typically achieved through significant effort and expense. Too often, the approaches used lead to interoperability that is specific to the targeted systems (sometimes called “point-topoint interoperability”) and that does not facilitate extension to other systems. Achieving large-scale and consistent interoperation among systems will require a consistently applied set of management, constructive, and operational practices that support the addition of new and upgraded systems to a growing interoperability web. Improvements in technology alone will not be sufficient. There must be parallel improvements in the ways that current and future interoperability needs are identified, and how organizations pursue interoperability. Applications of these models is made in chapters 5-7.

42


References 1. John Dunlop, Demessie Girma, James Irvine, “Digital Mobile Communications and the TETRA System”, Wiley www.tetramou.com www.nokia.com/tetra Doug Gray, “TETRA: The Advocate’s Handbook” ISBN 0-9544651-0-5 Roddy, Dennis, Satellite Communications, 2nd Edition, New York: McGrawHill, 1989:424. 6. Edwin Morris et.al. System of Systems Interoperability (SOSI) Final Report Carnegie Mellon Software Engineering Institute, ESC-TR-2004-004

2. 3. 4. 5.

3 TETRA Providing an Acceptable Security System Solution

Peter Stavroulakis

3.1 Introduction Having examined the characteristics of the general class of PMR systems and the requirements for the design of autonomous systems with maximal security, we conclude that TETRA is a good candidate for satisfying in a technically sound way these requirements. Of course, certain improvements in standards and applications will have to take place as we propose in chapters 4-7. In order to prove our point, a comparison of TETRA with its closed competitors is in order. This chapter, a substantial part of which is adapted from [1], the differences between TETRA and a GSM solution will be analysed in three dimensions in the hope to exhibit the clear and perhaps unique advantages of TETRA for security applications as will be shown in the subsequent chapters. The first dimension compares the applicable ETSI specifications and points out which functions are available according to the standard. Proprietary solutions will not be discussed on that level. The second dimension, a technical analysis, discusses how the end users and the operators perceive the differences between the network-solutions. Since it is possible technically to provide TETRA capabilities using GSM, an economic analysis focusing on the cost of the two alternative solutions, including capital and operational costs for the network infrastructure and end-user terminals which constitutes the third dimension, will be touched upon only briefly, because the applications of secure systems do not depend as much on economic terms but on their technical feasibility and the existence of international standards and on their ability to satisfy certain predefined requirements.

44


The methodical approach will be an analytic hierarchy process. Since there is a very strong correlation between the first and the second dimension, the results from the comparison of the air interface specifications will be used as an input for the technical analysis. Therefore, only the results from the technical and economic analysis will be used to derive final conclusions.

3.2 Hierarchical analysis The picture below shows the structure of the hierarchical analysis.

Fig. 1. Structure of analysis

Three different network solutions will be discussed: • A new, independent TETRA network • A GSM ASCI network which is based an existing GSM network • A GSM ASCI overlay network which is based on an existing GSM network where only part of the cells are upgraded to support group call functionalities. This approach assumes a certain cell overlapping in the GSM network and the comparison will be based on the following features and capabilities. 3.2.1 Air interface specifications Group call functionality has an effect on nearly all elements in the network. For TETRA only the air interface is specified; the core network, including, for example, the signalling between base station and exchange, is manufacturer specific. In contrast, the GSM standard specifies also the interfaces between the network elements. The analysis focuses on the air interface specifications since they can be compared directly between the

3.2 Hierarchical analysis

45

two standards. Impacts on the core network will be mentioned as far as they are relevant. 3.2.2 GSM ASCI

The ASCI (Advanced Speech Call Item) have been originally developed by ETSI and UIC (International Union of Railways). Therefore, the functional specifications are targeted firstly on railway communications. The ASCI features are part of GSM phase 2+ and consist of the following three items • Voice Group Call Service (VGCS) — A teleservice • Voice Broadcast Service (VBS) — A teleservice • Enhanced multi-level precedence and pre-emption service (eMLPP)-A supplementary service None of the ASCI features can be used with phase 1 or 2 mobiles. This means that dedicated terminals are required. Fallback and direct mode are both not defined in any GSM standard. An implementation would require proprietary solutions. 3.2.3 Enhanced Multi-Level Precedence and Pre-emption service (eMLPP)

The service consists of two parts — precedence and pre-emption. Precedence allows assigning priority levels to calls in combination with fast call set-up. If a higher priority call is set-up when all resources are in use, preemption allows seizing of lower priority resources. There are 7 different priority levels. The two highest ones are reserved for network internal use (e.g. for specific broadcast calls) and are only available for calls within one MSC area. Additionally three different classes of set-up time performance levels are defined. For each user in the network, the maximum precedence level may be defined. If the user does not use eMLPP services, the network uses a default priority. Calls of highest priority and fast call set-up do require neither authentication nor encryption on the radio link. Authentication and encryption may be postponed or omitted. The information about the different precedence levels is stored in the SIM of the mobiles and in the HLRNLR of the network. Implementation of the feature requires new features and parameters in MS, BTS, BSC and MSC. User data storage is needed in SIM and HLRNLR.

46


3.2.4 Voice Group Call Service (VGCS)

VGCS gives the possibility to establish group calls in a GSM cellular network [1]. A group call has basically three different kinds of participants: the talking subscriber, the listening subscribers and the dispatchers. Talking and listening subscribers can only participate in the group call, if they are within a predefined group area and if the group ID is stored to their SlM. Dispatchers can be located anywhere in the network and they are identified by MSISDN or ISDN. The maximum amount of dispatchers in a call is five. This restriction is caused by the definition of the conference call. The service requires a new network element, the Group Call Register (GCR). It contains group details such as group area and dispatchers. The interface between MSC and GCR is not specified and therefore vendor specific. The group IDs are stored in the mobiles’ SIM; updating of subscriber data over the air interface is not considered in the current specifications. This means that DGNA is not supported. For the calling subscriber, a standard call set-up procedure, depending on the priority, is done. The BSC allocates resources and then invites the group members to the call using the new logical channel Notification Channel (NCH). The NCH sends the information for the whole duration of the call. This allows group members which are in the beginning of the call outside the group area, to join the call (late entry). The network may pre-empt resources of lower priority. This is possible for emergency calls based on the specifications for eMLPP which are described above. Only one mobile subscriber can talk at any moment, the other participants can listen to the common downlink channel, which means that if several subscribers are located in the same cell, they will listen to the same channel. By pressing the PTT, a listening subscriber can request a speech item. Speech-items are allocated on first come first serve basis without queuing. The talking subscriber always reserves a separate traffic channel on the uplink. The dispatchers can talk at all times and their speech is connected to the common downlink channel. For the talking subscriber, standard handover procedures can be used within the group area. Listening subscribers have to initiate the handover themselves, since the system does not have any information about the users in the call. Seamless handover is not supported; idle mode cell reselection has to be used. A dispatcher uses standard handover procedures. The described group functions are related to speech only. SMS to group numbers are not supported. A listening subscriber cannot receive

3.3 TETRA

47

any signalling while in the call, which means that e.g. sending and receiving of SMS during a call is not possible. The calling subscriber or a dispatcher can terminate the call. Termination through inactivity after expiring of a timer is also possible. For the calling subscriber, authentication and encryption are optional. They may be omitted or postponed when using fast call set-up. For listening subscribers, authentication is not possible but encryption is optional. Like for eMLPP, modifications in all major network elements are required in order to support this feature. These are MS, BTS, BSC, MSC, VLR/HLR and SIM. Additionally the implementation of GCR is needed. 3.2.5 Voice Broadcast Service (VBS)

VBS consists of the same functionalities as VGCS with the difference that the speech is unidirectional. The calling subscriber may be a mobile user or a dispatcher. No uplink functionality is required for the listening subscribers. The network requirements are similar as for VGCS. 3.3 TETRA The group call functionalities are defined in the ETSI specifications Terrestrial Trunked Radio (TETRA), Voice plus Data (V+D), Part 2: Air Interface (AI) [4]. These specifications also include the signalling for fallback and direct mode, however, they will not be discussed since a comparison to GSM is not possible. Every TETRA subscriber, mobile or dispatcher, is identified by an ITSI (Individual TETRA Subscriber Identity). A group is identified by a GTSI (Group TETRA Subscriber Identity). The ITSI/GTSI consists of a country code, a network code and the subscriber identity. A mobile may be the member of different groups at the same time and it sends its membershipinformation to the Switching Infrastructure (SwMl) upon registration. The group information, like group area and members, is stored in the SwMl. The standard specifies updating of group information to the mobiles over the air interface (DGNA), for example, by authorized dispatchers. When initiating a call, the subscriber sends the GTSI on the MCCH (Main Control Channel) to the SwMI. The SwMI allocates one traffic channel on every site within the group area and invites the mobiles to the call. Optionally it may reserve resources only on those sites, where members are located (shifting group call area). Priority queues are used to allocate resources. Pre-emptive services are also supported.

48


Group calls are always established as semi-duplex calls for all subscribers. By pressing the PTT, a listening subscriber can request a speech item. Speech-items are allocated using a priority queue. The talking subscriber uses the already allocated TCH (traffic channel) on the uplink. In contrast to GSM, the mobile always initiates handovers in TETRA. The standard describes three different types of declared handovers. Seamless handover is supported for the talking subscriber. The listening subscriber can make an undeclared handover, which means that the mobile needs to move to the control channel of the new site from where it will be commanded to a traffic channel. If a subscriber moves to a cell within the group area where no TCH is active for the call, the SwMI has to allocate a TCH to the group call for that subscriber. Since quasi transmission trunking is commonly implemented, the call will be terminated after expiring of a certain hang time. The standard also allows termination through the calling subscriber or a dispatcher. Authentication is done during registration and roaming. The standard supports encryption with static or dynamic keys. 3.3.1 Comparison of specified features Since TETRA has been especially developed for group calls, it fulfils most of the requirements. Some of the limitations in the table below are stated as “unlimited”. They reflect the standard, however effective limitations are, manufacturer specific. Shifting group call area is basically possible for both technologies, TETRA and GSM ASCI. The air-interface specifications do not mention the implementation, since resource allocations are part of the core network functionalities. Several manufacturers have implemented the feature for TETRA; however, for GSM it does not exist. An implementation would require major modifications in the core network software, which is not feasible for economic reasons.[1] Table 3.1. Summary of technical comparison Resource queuing priorities Resource pre-emption Speech item priorities Speech item pre-emption Late entry Maximum group size Maximum amount of dispatchers in group

TETRA Yes Yes Yes Yes Yes Unlimited Unlimited

GSM / ASCI Yes Yes No No Yes 1024 5

3.3 TETRA Maximum amount of sites in group call Amount of groups per subscriber Fixed group area Shifting group area Used traffic channels for a group call on N sites with X mobile dispatchers Authentication of talking subscriber Authentication of listening subscriber Seamless handover for talking subscriber Fast handover for listening subscriber

Unlimited Unlimited Yes Yes

Unlimited 50 Yes No

N Yes Yes Yes ~1s

N+X+1 Optional No Yes No

49

All above-mentioned features are input-information for the functional and operational comparison in the next chapter. Therefore, table 1 has not been taken into account in the hierarchical analysis process. 3.3.2 Technical analysis The section below analyses, how the end-users and the network operators perceive the functionalities described in the previous section. It also discusses, which are the restrictions caused by certain solutions and how strong impact they have in practical situations. The section is divided into three parts: network functions, capacity and security. Network functions

The network functions reflect which services are available for the end users including dispatchers and mobile users. Services, which have direct impact to capacity, will be discussed in a separate chapter. In contrast to commercial networks the functions also include network management tasks like the creating of groups. This so called tactical management allows dispatchers, for example, to create new groups for an upcoming mission (DGNA). Group size

The TETRA specifications do not mention any group size restriction. Limitations are vendor specific. In case of Nokia SwMI, the amount of radio subscribers in a group is not limited and the maximum amount of dispatchers in a group call is 30. GSM ASCI limits the amount of mobile subscribers in a group to 1024 and the amount of dispatchers to 5. The first limitation may be significant for broadcast group calls to large organisations. The second limitation is

50


especially critical if different organisations are involved in a group call. Cross-organisational communication is important, for example, during a major incident; see also below (DGNA). Dynamic Group Number Allocation (DGNA)

A dispatcher has to be able to allocate mobiles to temporary groups. This allows allocating resources for a mission, even if the mobile users belong to different organisations. For GSM ASCI, there are different solutions to bypass the nonexistence of DGNA: • A set of pre-programmed groups could be used for special missions. However, this approach is hardly applicable for cross-organisational communication, since it highly compromises the numbering flexibility. • A conference bridge could be used to combine groups of existing organisations [18]. This would reduce the amount of included dispatchers because of the limitation of S participants in a conference call. • Proprietary solutions for over the air programming of the group data via SMS have been considered [1]. This seems to be the only acceptable solution for end-users. The operators and end-user organisations should be aware of the fact that proprietary solutions often have negative influence on interoperability. Furthermore, the lack of competition typically causes higher equipment prices. Short data messaging

The sending of short data messages to groups is a very effective way of informing the members of a group during an incident. This feature is supported by TETRA but not by GSM ASCI. The sending of individual messages, as bypass, is significantly slower and causes additional load on the signalling channel. This increases the risk for control channel congestion during peak loads like emergency situations. Additionally, it is not possible to send individual SDS to mobiles which are engaged in a group call. This solution is not acceptable for the use in critical situations. Call priorities

Queuing and pre-emptive priorities are specified for TETRA as well as for GSM ASCI. The Difference is that in GSM priorities are associated to

3.3 TETRA

51

subscribers. In TETRA different priorities can be given to subscribers or groups which lead to higher flexibility. Furthermore, TETRA supports more priority classes than GSM ASCI. Main difference, however, is the lack of speech item priorities in GSM. In emergency situations it is important that a group leader can get a speech item and force the others to listen in order to achieve organised communication. Priority scanning

No signalling for priority scanning is specified for GSM ASCI. If a member is engaged in a group call, she or he is not able to receive any other signalling. This restriction may be very critical for PSS users, where certain high priority group communications need to be available even if some of the members are engaged in another group call. DMO and base station fallback

In case the mobiles are out of network coverage or the connection between exchange and base station is down, for example, due to a major accident, GSM based mobiles are not able to communicate. No base stations or mobiles supporting these functionalities are available on the market for the time being. The TETRA standard specifies both functionalities. Network capacity

The required network capacity depends on different factors, like cell size, support of shifting area group call and the end-user requirement for call set-up times. The following subchapter discusses the impact of these factors on the amount of required traffic channels and bandwidth. Cell sizes

Average TETRA cells are remarkably larger than GSM cells. Firstly, TETRA uses typically a frequency of 400MHz, while GSM uses 900 or 1800MHz. The propagation losses are theoretically proportional to the square of the frequency [12]. Secondly, commercial networks are typically capacity driven and PSS networks with less users are coverage driven. This means that the population density usually determines cell sizes in GSM. The planned TETRA network in Germany consists of roughly 3000 cells. In contrast to that, the existing GSM network from Vodafone has 38100 cells using 15700 base stations. Assuming that most of the base

52


stations have either 1 or 3 cells, we get 11200 three-cell base stations and 4500 one-cell base stations. Overall we get an average relation of GSM:TETRA cells of 12.7:1. Due to the high capacity requirements in densely populated areas, the relation is somewhat higher in urban than in rural regions. Taking into account the fact that there is a relatively high cell overlapping in the existing GSM networks, it would be basically possible to use only part of the cells for group calls. In urban areas, the network often consists of an overlay network providing coverage over a large area and micro cells improving location probability and providing capacity in densely populated areas. Building an overlay network by using only macrocells for the group calls significantly reduces network costs because not all base stations need to be upgraded to support ASCI functionalities. Additionally we will see later that it reduces the generated load in the network. The main drawback, however, is a lower location probability which leads to a lower quality of service. The table below shows the assumptions for the calculations in the next chapters. The average cell sizes for the GSM overlay solution are assumed to be double compared to the commercial GSM network. Table 3.2. Cell sizes and bandwidths TETRA GSM ASCI GSM ASCI overlay Average cell area (km2)

120

10

Frequency re-use factor

19

12

12

Bandwidth per carrier (kHz) 25

200

200

Bandwidth per channel (kHz) 6.25

25

25

20

Traffic modelling of group calls

Models are commonly used to estimate the total traffic in a telecommunication system. In commercial networks the input data is subscriber density, area coverage and busy hour traffic per subscriber [1]. The call intervals as well as the call durations are assumed to be exponentially distributed (socalled Poisson traffic). In this case, the required amount of traffic channels can be determined by using the Erlang C-formula which is shown below.

3.3 TETRA

53

Group calls used in PSS networks, however, have some special characteristics: • The amount of users is known which means that the user density is derived from the total amount of subscribers and not vice versa as in commercial networks. • In contrast to commercial networks, traffic is assumed to be constant over a longer period of time. The traffic growth is limited due to the constant amount of users. Network capacity seldom needs to be added. Network expansions are mainly for improving coverage. • Semi-duplex group calls reserve one traffic channel one each site which is activated during the group call. Therefore, a traffic model for group calls typically includes the amount of cells activated in an average call. The assumption is that if a cell generates traffic to other cells due to group calls, it will also have to take traffic from other cells in the same proportion. • The variance of the call intervals is higher and calls are of short duration since communication is used for tactical operation. During incidents, the capacity is significantly higher than during normal times, which leads to a bustier traffic distribution than Poisson traffic. ETSI recommends evenly distributed call durations and exponentially distributed call intervals for PSS traffic models [6]. In practice, however, Erlang C formula, assuming exponentially distributed call durations and intervals, is commonly used [1]. Erlang C formula: The Erlang C formula assumes a queuing system and determines the probability that a call needs to wait longer than a certain queuing period. This probability is determined by using the two formulas which are described below. P describes the probability that a call is delayed (i.e. it needs to queue for resources):

PO =

AX ⎡ ⎛ N − A ⎞ N −1 A X ⎤ A X + ⎢ N !•⎜ ⎟• ∑ ⎥ ⎣ ⎝ N ⎠ X =0 X ! ⎦

(3.1a)

N = Number of available communication channels A = Traffic intensity in Erlangs (λ*H), where λ = call interval H = Mean holding time (i.e. average call duration) in seconds The value PΤ describes the probability for exceeding a certain queuing time T. This probability is also called grade of service (GoS):

54

3 TETRA Providing an Acceptable Security System Solution ⎛ ( N − A )•T ⎞ ⎜ ⎟ H ⎠

P(W > T ) = PT = P0 • e⎝

(3.1b)

W = Call waiting time in a FIFO queue. T = A given queuing period in seconds A typical value in PSS networks is a probability of 5% that the queuing time exceeds 5 seconds. The described GoS is correct for a single site. In order to get the probability that all sites are included into a call after a certain waiting time, the value is depending on the amount of sites in the group call. Assuming that PT is similar for all sites, we get the total PT according to the following formula: PT(total) = 1-(1-PT)n n = Amount of sites in the group call Because group calls include different amount of sites, this effect is usually neglected in the calculations and only the site GoS is calculated. Generated traffic: Taking into consideration the assumption above, the generated traffic per cell is: A=U·λ·H·n (3.2) U = Amount of users per site λ = Busy hour call attempts per subscriber H = Average call duration (λ · H = traffic per user) n = Amount of cells included in a group call In a similar way we can also define the traffic generated on a site by one single group: A=λ·H·G (3.3) G = Amount of active subscribers in the group The amount of cells included in a group call is depending on the group size, the location and the moving behaviour of the group members as well as the average cell size. These parameters differ between the main user groups, police, fire brigades and ambulances. Below is a short description of their typical call behaviour. Police

In normal situations, the average group size for traffic police is about 50 and the members are spread over the whole area or district. This group is used for informative purposes. During a mission, a group size is typically about 5 and the members are located in a certain, smaller area. The

3.3 TETRA

55

moving behaviour but also the size may be very different depending on the mission type. Fire brigades & ambulance

During a mission, the group size is between 5 and 20 members which are located in a certain area. Members are typically moving only within small distances except when going to or leaving from places of interest. The so-called user profile for traffic modelling uses a weighted average of the above described values between the user groups. The numbers below are based on traffic in existing TETRA networks [1]. The amount of active sites per call describes the amount of sites where at least one active subscriber is located. The numbers have been estimated using an even distribution of the subscribers in one forth of the group area, because in case of an incident, group members are typically located in a relatively small area. Table 3.3. PSS user profile TETRA GSM ASCI GSM ASCI overlay Mobile originated traffic (mErl) Group size (users) Group call area (cells for 400 km2) Active sites per call

7.3 10 4 2

7.3 10 40 6

7.3 10 20 5

Call set-up time and open channel

The call set-up time requirement for PSS users is 0.5s. Currently this is only achievable with TETRA and transmission trunking can be used for channel allocation. In GSM it would be basically possible to reduce the call set-up time to about one second, but this would require major changes in the core network software and topology. In practice this is for economic reasons not feasible in commercial networks. Therefore in GSM a traffic channel needs to be allocated for the whole duration of a communication (message trunking). If the call intervals are not predictable at all, a channel needs to be activated for a group all the time (open channel). The amount of simultaneously active groups, which can be served per site, is significantly higher when using transmission trunking. The table below has been calculated using formulas (3.1) and (3.3) with a GoS of 5% for exceeding 5 seconds queuing time and an average call duration of 12

56


seconds. For example, a 2-carrier TETRA base station can serve about 8 times more groups than a single carrier GSM base station even though both have the same amount of traffic channels. Amount of served groups per site 60

57

50 45

40 34 30 24 20 14 10

6 1

1

7

6

5

4

3

2

0 1

2

3

4

5

6

7

Available traffic channels Open channel

Transmission trunking

Fig. 2. Transmission trunking versus open channel operation

Shifting area group call

The support of shifting area group calls has a significant effect on the amount of traffic channels involved in a group call. Using the subscriber profile above, the amount of traffic channels which needs to be allocated for a group call, would be 2 for TETRA and 6 for GSM ASCI in case shifting group call area is supported. However, if this feature is not supported and high moving interest of the subscribers is assumed, all cells in the group area need to be activated. This would be 4 for TETRA and 40 for GSM. Taking into account that shifting area can only be realized for TETRA, the relation of channel occupations for a group call between TETRA and GSM is 1:20 in this example. It may be possible [1] that a group call activate all cells where members are located plus the adjacent cells. It is obvious that this solution could only be used for user groups with low moving interest or if transmission trunking is supported. Otherwise, in case of message trunking, group members may lose the communication to the group when moving to other sites during the relatively long call duration. Additionally the small cell sizes in GSM increase the probability for cell changes during a call.

3.3 TETRA

57

Required capacity on the radio network

Since the cell sizes are very different for TETRA and GSM it is not representative to observe the amount of used traffic channels. The required bandwidth or the amount of transceivers is more representative. The bandwidth can be either expressed in terms of radio channels or frequency. The required bandwidth depends on the carrier capacity per site, the bandwidth per channel and the frequency re-use factor R. If the amount of cells in the network is smaller than the reuse-factor, then the amount of cells has to be used as R in the formula below. With a given carrier capacity C per site, the needed bandwidth in the network would be: TETRA:

B = C · 25 kHz · R

GSM:

B = C · 200 kHz · R

We are considering two cases: a regional network with an area of 900 km and 400 users representing a medium size city with rural surroundings and a Germany-wide network with an area of 357 021 km2 and 529 000 users. In both cases, the average group calf area is 400 km2 and a group consists of 10 members. The assumption is that all users are active during busy hours. This might seem like an overestimation of the traffic, however, we will see that the calculated capacity for TETRA in table 5 is even slightly less than the recommendations for the Germany-wide network. Since GSM does not support shifting area group calls, nor does it fulfil the required group call setup times, open channels have to be used for some groups. The following assumptions have been taken: • One radio channel on each site belonging to the group call area will be activated during a group communication. • 25% of the groups use an open channel for the communication, which means that a traffic channel is allocated all the time for the group. • 75% of the groups use transmission trunking, which means that these groups have to accept longer call set-up times. • The additional load generated by the talking subscriber and by mobile dispatchers has been neglected. The channel holding time for quasi-transmission trunking has been neglected in both cases, GSM and TETRA. The generated traffic and required capacities per cell has been calculated using formulas 1 and 2 for groups using transmission trunking. The selected grade of service (G0S) corresponds to a probability of less than 5% that the channel queuing time would exceed 5s. In case of open channel, one Erlang is required on each site within the group call area during busy hours. Since GSM ASCI is built on top of an existing network, the additionally required capacity has been determined. As reference, the Vodafone

58


network has been taken, which consists of 38100 cells and 93600 transceivers [1] and has in average 18 traffic channels per cell. We assume that the capacity in the network is optimised on all sites, which means that 18 traffic channels carry up to 14.1 Erlang traffic during busy hours at the given GoS. This value has been added to the generated traffic through group communication and the sum has been used to determine the required amount of traffic channels. The traffic on the control channel has not been analysed and it has been assumed that no additional control channels are required for GSM ASCI. The average amount of carriers is a statistical value which therefore may be any decimal number. In case of TETRA, each site has one control channel for signalling purposes. Table 3.4. Required radio network capacity for a small network[1] Network size Technology

900 km2 and 400 users TETRA GSM ASCI GSM ASCI overlay 8 90 45

Amount of cells Average amount of users per 50 cell Trunked traffic per cell 0.73 Open channel traffic per cell 0 Total traffic per cell in Erl 0.73 Required traffic channels per 3 cell (GoS = 5% at 5s) Required amount of carriers per 1 cell in average Required amount of transceivers 8 in total Required bandwidth in carriers 8 Required bandwidth in kHz 200

4.4

8.9

0.96 4.44 5.41

0.97 4.44 5.42

7

7

0.875

0.875

79

40

11 2200

11 2200

In this example, the required bandwidth for TETRA is 11 times less than for GSM. Interesting is that the traffic per cell and the bandwidth requirement are the same for both GSM solutions. However, the amount of transceivers in case of the overlay network solution is only half because of the double average cell size. Table 3.5. Required radio network capacity for a large network [1] Network size Technology Amount of cells

357021 km2 and 529000 users TETRA GSM ASCI 2976 35703

GSM ASCI overlay 17852

3.3 TETRA Average amount of users 177.8 per cell Trunked traffic per cell 2.60 Open channel traffic per 0 cell Total traffic per cell in Erl 2.60 Required traffic channels 6 per cell (GoS = 5% at 5s) Required amount of carri1.75 ers per cell in average Required amount of trans5208 ceivers in total Required bandwidth in 34 carriers Required bandwidth in 850 kHz

14.8

29.6

3.24

3.24

14.82

14.82

18.06

18.06

19

19

2.375

2.375

84795

42399

29

29

5800

5800

59

It is clear that the amount of sites is inverse proportional to the cell size. Together with the required bandwidth, this determines the total amount of transceivers in a network. Therefore, a TETRA solution requires significantly less carriers than GSM, even when regarding an overlay solution. The amount of transceivers does not only determine the size of the base stations, but also the amount of required transmission lines between base stations and exchanges or base station controllers. This value has therefore a major impact on the operational costs of the network. The ratio of required bandwidths of GSM: TETRA is about 7:1 for the countrywide network. The reason that this ratio is smaller than for the regional network comes from the fact that GSM has a lower frequency reuse factor than TETRA and TETRA does not need to reuse frequencies in the small network. For the GSM solutions, the size of the chosen group call area and the user density are directly proportional to the required bandwidth. In case of TETRA, however, the size of the group call area has no influence to the capacity since shifting area can be used. The graph below shows the required bandwidth in dependence of the average group call area. Input data are the same as for the calculation of the countrywide network above where 25% of the groups use open channel communication. For TETRA, the required bandwidth is constant at 850 kHz for a constant amount of users. If the group call area is 20 km2 or less, the required bandwidth for GSM is 600 kHz. For larger areas, the required bandwidth grows more or less proportionally to the group call area and reaches significantly higher values than TETRA.

60


Fig. 3. Influence of group call area to bandwidth

As discussed earlier, the location of the groups is very seldom predictable and therefore, smaller group call areas can have a major impact on the quality of service perceived by the end users. For example, traffic police units move in very wide areas and a restriction of the group call area is in practical situations not acceptable. Therefore, a direct comparison between the two solutions is very difficult and one has to keep in mind that a network supporting shifting group call area always offers better quality of service since it is not realistic to use extremely large fixed group call areas. For further discussions, especially economic considerations, a group call area of 400 km2 has been chosen. It is obvious that keeping a channel open all the time wastes radio resources. The graph below shows, how the bandwidth requirement grows in the GSM network when groups are using open channel communication instead of transmission trunking. The calculations assume a group call area of 400 km2. TETRA can offer fast call set-up times even if transmission trunking is used and therefore the bandwidth requirement is constant at 850 kHz. For GSM ASCI open channel communication has to be used for groups where fast call setups are required. The required bandwidth grows

3.3 TETRA

61

nearly linearly from 1.6 MHz if all groups use transmission trunking, to 18 MHz if all groups use open channels.

Fig. 4. Influence of open channel communication to bandwidth

For further discussions, it is assumed that 25% of the users require fast call set-ups and therefore use open channel communication in GSM. Network security

Network security is mainly important for organisations where the communication is secret. The network should ensure that one not belonging to the group is unable to listen to the communication but also unable to disturb the communication. It should also be impossible to follow a certain subscriber by tracing or recording the signalling on the control channel. Authentication

Authentication ensures that only mobiles with a valid key are able to use the network. In TETRA, authentication is done during the registration. The network rejects mobiles which return a wrong authentication key. In GSM, authentication is done during call set up and may be omitted during fast call set-up, Listening members in a group call are not authenticated in GSM ASCI since there is no uplink signalling during the call. This means that every mobile which has the group ID programmed to its SIM card, is able to listen to a group call. This is a high security threat for many PSS customers. In TETRA networks it is possible that the mobile authenticates the network. This mutual authentication enables the mobile to detect fake base

62


stations, thus it will not register to base stations not belonging to the network. Pseudo mutual authentication is also possible by using dynamic authentication keys. In such case it is not possible for a fake network to authenticate a subscriber, since a new key is used for each registration. Air Interface Encryption (AIE)

AIE encrypts all signalling and call information on the radio path. Besides of the speech, it also encrypts the identities of the mobiles and the data messages on the control channel. This means that it is not possible to trace a mobile, for example, by following the signalling on the control channel. AIE is in use for both standards, GSM and TETRA. The TETRA algorithm uses longer keys and is supposed to be more secure than the one in GSM. End-to-end encryption (e2ee)

E2ee encrypts speech and data between the end-points of the communication. Encryption and decryption are done in the end-terminals. The network infrastructure offers a transparent transport layer which is supported by both technologies. Dynamic keys may be delivered to the terminals using SDS/SMS. Even though, e2ee completely protects against eavesdropping, it cannot encrypt signalling information. E2ee encrypts all information on the traffic channel but not on the control channel. Therefore e2ee has to be used in combination with AIE. Summary of comparison on technical level

The analysis has been split into three areas: network functionalities, capacity and security. Each of the areas will be summarized separately in order to keep high transparency. Since not all the features have the same importance, different weighting-factors have been used: • Critical features: 2 • Default weight: 1 • Minor features: 0.5 The fulfilment of the requirements has been scored as follows: • • • • •

Completely fulfilled: 100% Minor functionality missing: 75% Major restrictions in functionality: 50% Only a bypass solution existing: 25% Feature not supported at all: 0%

3.3 TETRA

63

Since the priorities of features are different for the various user groups, different weighting and scoring can be applied. The chosen values are interpretations from the related literature and practical experience. The different functionalities and features (arguments) have been summarised in tables which all have the following structure:

Argument A Argument B Argument C Argument D Normalised sum of weighted grades

Weight WA WB WC WD

Technology 1 A1 B1 C1 D1

Grade1 Grade2



Grade3

The normalised sum of weighted grades (Grade1) has been determined according to the following formula:[1]

⎛ ⎞ ⎜ ⎟ 1 ⎜ W A ⋅ A1 WB ⋅ B1 WC ⋅ C1 WD ⋅ D1 ⎟ Grade1 = D ⋅⎜ 3 + 3 + 3 + 3 ⎟ W A B C D ⎜ ∑ ∑ ∑ ∑ x ⎜ ∑ y y y y ⎟ ⎟ x= A y =1 y =1 y =1 ⎝ y =1 ⎠ Grade2 and Grade3 are calculated in a similar way. Regarding the group call functionalities of how they are perceived by end-user, GSM ASCI lacks of major functionalities which may be crucial for PSS users. Most of the missing features cannot be even compensated using proprietary features since they are directly related to the air interface signalling. For example, speech item priorities cannot be implemented unless specified by the standard. Another major drawback of GSM ASCI is that no signalling is possible to listening members in a group call. The long GSM ASCl call set-up times cause that open channel communication has to be used for certain user groups. This means that mobiles are not able to receive any individual calls or SDS as long as the group radio channel is open. Table 3.6. Network functions

Group size limitations

Weight TETRA GSM ASCI GSM ASCI overlay 1 100% 50% 50%

64


DGNA Group messaging (SDS) Call priorities Speech item priorities Priority scanning Late entry Fast handover Radio coverage Direct mode Base station fallback Normalised sum of weighted grades

1 0.5 2 2 1 1 1 1 1 1

100% 100% 100% 100% 100% 100% 75% 75% 100% 100%

25% 25% 75% 0% 0% 100% 50% 75% 0% 0%

25% 25% 75% 0% 0% 100% 50% 75% 0% 0%

0.675

0.168

0.158

The long call set-up times and the non-existence of shifting area group calls have major impact on the required air interface and core network capacity. Since GSM ASCI has been first of all developed for railways, shifting area has not been a crucial argument during the development of the standard. For PSS customers, where the location of the users is not known in advance and cell changes must be possible, shifting area is a critical requirement. The row traffic channel usage takes into account that the speaking user as well as all mobile dispatchers require an own traffic channel in GSM. Table 3.7. Network capacity

Cell size Frequency reuse Bandwidth per channel Call set-up time Shifting group call area Traffic channel usage Normalised sum of weighted grades

Weight TETRA GSM ASCI GSM ASCI overlay 0.5 100% 50% 75% 1 50% 100% 100% 1 100% 50% 50% 2 100% 25% 25% 2 100% 25% 25% 0.5 100% 75% 75% 0.541

0.225

0.233

Main drawback of GSM ASCI related to security is the fact that listening members in a group call are not authenticated and authentication may be skipped or postponed if fast call set-up is used. The importance of the security arguments differs much among the user groups. Rescue forces, like ambulance or fire brigades, typically do not need 100% protection against eavesdropping, however, for police forces proper authentication and encryption is a must.

3.3 TETRA

65

Table 3.8. Network security

Authentication AIE E2EE Normalised sum of weighted grades

Weight TETRA GSM ASCI GSM ASCI overlay 1 100% 50% 50% 2 100% 50% 50% 0.5 100% 100% 100% 0.476

0.262

0.262

In all three areas, GSM ASCI shows clearly lower performance than TETRA. The summary of the technical analysis is a weighted sum of the discussed areas. Network functions and security are weighted with 40% and network capacity with only 20% since its influence is mainly on economic level, where it will be considered once more. Table 3.9. Summary of technical analysis Weight 40% Network functions 20% Network capacity 40% Network security Normalised sum of weighted grades

TETRA 0.675 0.541 0.476

GSM ASCI 0.168 0.225 0.262

GSM ASCI overlay 0.158 0.233 0.262

0.569

0.217

0.214

The graphical representation shows that the advantage of TETRA is similar in all three areas, which means that using different weighting has hardly any influence to the result of the technical analysis.

66


Proportional points

Summary of technical analysis

TETRA Network functions

GSM ASCI overlay

Network capacity

GSM ASCI Network security

Fig. 5. Summary of technical analysis

A cost analysis based on the above technical comparison shows [1], that as far as which system provides a cost saving solution, by taking into consideration both Operating Expenses and Capital Expenses TETRA, provides a distinct advantage. This advantage is maintained even in the case when risk elements are included which make TETRA more vulnerable. As a matter of fact even combining economic results with technical for an overall comparison, TETRA wins by far. Having these results as the basis of our optimism, in the following chapters, we present innovative and unique implementations of TETRA, in an effort to show that as TETRA improves, it can be used as the core of unified secure systems to handle WLANS, AD-HOC. An innovative improvement is proposed in the next chapter for the purpose of achieving an optimal channel assignment scheme.

References 1. Simon Riesen, The Usage of Mainstream Technologies for Public safety and Security Networks, Master’s Thesis, 2003, Department of Electrical and Communications Engineering, Helsinki University of Technology

4 Channel Assignment and Multiple Access in Trunking Radio Systems [1]

Peter Stavroulakis, Kostas Ioannou, Ioannis Panoutsopoulos

4.1 Channel Assignment Techniques [1]

4.1.1 Introduction Technological advances and rapid development of hand-held wireless terminals have facilitated the rapid growth of mobile security networks and mobile computing. Taking ergonomics and economics factors into account, and considering the new trends in the telecommunications industry to provide ubiquitous information access, the population of mobile security users in TETRA Networks will continue to grow at a tremendous rate. Another important developing phenomenon is the shift of many applications to multimedia platforms in order to present information more effectively. The tremendous growth of the wireless/mobile users in mobile security networks coupled with the bandwidth requirements of multimedia applications requires efficient reuse of the scarce radio spectrum allocated to wireless/mobile communications. Efficient use of radio spectrum is also important from a cost-of-service point of view, where the number of base stations required to service a given geographical area is an important factor. A reduction in the number of base stations and hence a reduction in the cost-of-service can be achieved by more efficient reuse of the radio spectrum. The basic prohibiting factor in radio spectrum reuse is interference caused by the environment or other mobiles. Interference can be reduced by deploying efficient radio subsystems and by making use of channel assignment techniques.

68


In the radio and transmission subsystems of TETRA Network, techniques such as deployment of time and space diversity systems, use of low noise filters and efficient equalizers, and deployment of efficient modulation schemes can be used to suppress interference and to extract the desired signal. However, co-channel interference caused by frequency reuse is the most restraining factor on the overall system capacity in the wireless networks and the main idea behind channel assignment algorithms is to make use of radio propagation path-loss [2, 3] characteristics in order to minimize the Carrier-to-interference ratio (CIR) and hence to increase the radio spectrum reuse efficiency. The focus of this chapter is thus to provide an overview of different channel assignment algorithms as they relate to TETRA Networks and compare them in terms of performance, flexibility, and complexity. We first start by giving an overview of the channel assignment problem in a cellular environment and we discuss the general idea behind major channel allocation schemes. Then we proceed to discuss different channel allocation schemes within each category and we follow with the development of an optimization technique in channel assignment. 4.1.2 Channel Allocation Schemes What Is Channel Allocation?

A given radio spectrum (or bandwidth) can be divided into a set of disjoint or non-interfering radio channels. All such channels can be used simultaneously while maintaining an acceptable received radio signal 1. In order to divide a given radio spectrum into such channels many techniques such as frequency division (FD), time division (TD), or code division (CD) can be used. In frequency division, the spectrum is divided into disjoint frequency bands, whereas in time division the channel separation is achieved by dividing the usage of the channel into disjoint time periods called time slots. In code division, the channel separation is achieved by using different modulation codes. Furthermore, more elaborate techniques can be designed to divide a radio spectrum into a set of disjoint channels based on the combination of the above techniques. For example, a combination of TD and FD can be used by dividing each frequency band of a FD scheme into time slots. The major driving factor in determining the number of channels with certain quality that can be used for a given wireless spectrum is the level of received signal quality that can be achieved in each channel.


69

Let Si(k) be denoted as the set (i) of wireless terminals that communicate to each other using the same channel k. By taking advantage of physical characteristics of the radio environment, the same channel k can be reused simultaneously by another set j if the members of set i and j are spaced apart sufficiently. All such sets which use the same channel are referred to as co-channel sets or simply co-channels. This is possible because due to propagation path-loss in the radio environment, the average power received from a transmitter at distance d is proportional to PT d-a where a is a number in the range of 3-5 depending on the physical environment and PT is the average transmitter power. For example, for an indoor environment with a = 3.5, the average power at a distance 2d is about 9% of the average power received at distance d. Thus by adjusting the transmitter power level and/or the distance between cochannels, a channel can be reused by a number of co-channels if the carrierto-interference ratio (CIR) in each co-channel is above the required value CIRmin. Here the carrier (C) represents the received signal power in a channel, and the interference (/) represents the sum of received signal powers of all co-channels. As an example, consider Figure 4.1 where a wireless station labeled R is at distance dj from a transmitter station labeled T using a narrowband radio channel. We refer to the radio channel used by T to communicate to R as the reference channel. In this Figure, we have also shown five other stations labeled 1, 2,… 5, which use the same channel as the reference channel to communicate to some other stations. Denoting the transmitted power of station i by Pi and the distance of station i from R by di, the average CIR at the reference station R is given by:

CIR =

Pt d t

−∞

∑i=1 Pi d i−∞ + N o 5

(4.1)

where No represents the environmental noise. To achieve a certain level of CIR at the reference station R, different methods can be used. For example, the distance between stations 1, 2, ..., 5 using the co-channel and the reference station R can be increased to reduce the co-channel interference level.

70


Fig. 4.1. Interference model

Many channel allocation schemes are based on this idea of physical separation. Another solution to reduce the CIR at R is to reduce the interfering powers transmitted from five interfering stations and/or to increase the desired signal’s power level Pt. This is the idea behind power control schemes. These two methods present the underlying concept for channel assignment algorithms in cellular systems. Each of these algorithms uses a different way to achieve a CIRmin at each mobile terminal by separating co-channels and/or by adjusting the transmitter power.

Different Channel Allocation Schemes

Channel allocation schemes for TETRA Networks can be divided into a number of different categories depending on the comparison basis. For example, when channel assignment algorithms are compared based on the manner in which co-channels are separated, they can be divided into Fixed Channel Allocation (FCA), Dynamic Channel Allocation (DCA), and Hybrid Channel Allocation (HCA). In Fixed Channel Allocation (FCA) schemes, the area is partitioned into a number of cells, and a number of channels are assigned to each cell according to some reuse pattern depending on the desired signal quality. FCA schemes are very simple, however, they do not adapt to changing traffic


71

conditions and user distribution. In order to overcome these deficiencies of FCA schemes, dynamic channel assignment (DCA) strategies have been introduced. In DCA, all channels are placed in a pool and they are assigned to new calls as needed such that the CIRmin criterion is satisfied. At the cost of higher complexity, DCA schemes provide flexibility and traffic adaptability. However, DCA strategies are in less efficient than FCA under high load conditions. To overcome this drawback at high load conditions, Hybrid Channel Assignment (HCA) techniques were designed by combining FCA and DCA schemes. Channel assignment schemes can be implemented in many different ways. For example, a channel can be assigned to a radio cell based on the coverage area of the radio cell and its adjacent cells such that the CIRmin is maintained with high probability in all radio cells. Channels could be also assigned by taking the local CIR measurements of the mobiles’ and base stations’ receiver into account. That is, instead of allocating a channel blindly to a cell based on worst case conditions (such as letting co-channels be located at the closest boundary), a channel can be allocated to a mobile based on its local CIR measurements [4, 5]. Channel assignment schemes can be implemented in centralized or distributed fashion. In the centralized schemes the channel is assigned by a central controller whereas in distributed schemes a channel is selected either by the local Base Station (BS) of the cell from which the call is initiated or selected autonomously by the mobile. In a system with cell based control each base station keeps information about the current available channels in its vicinity. Here the channel availability information is updated by exchange of status information between base stations. Finally, in autonomously organized distributed schemes, the mobile chooses a channel based on its local CIR measurements without the involvement of a central call assignment entity. Obviously, this scheme has a much lower complexity at the cost of lower efficiency. It is important to note that channel assignment based on local assignment can be done for both FCA and DCA schemes. Fixed Channel Allocation Scheme

In n the FCA strategy a set of nominal channels is permanently allocated to each cell for its exclusive use. Here a definite relationship is assumed between each channel and each cell, in accordance to co-channel reuse constraints [6-12]. The total number of available channels in the system C is divided into sets, and the minimum number of channel sets N required to serve the entire coverage area is related to the reuse distance s as follows [6, 12]:

72


N = (l/3)o2, for hexagonal cells (4.2) Here o is defined as DIRtt, where Ra is the radius of the cell and D is the physical distance between the two cell centers [5]. N can assume only the integer values 3, 4, 7, 9, ... as generally presented by the series, (i + j) 2 - ij, with i and/ being integers [5, 7]. Figures 2a and 2b give the allocation of channel sets to cells for N = 3 (o = 3) and N = 1 (a = 4.45), respectively. In the simple FCA strategy, the same number of nominal channels is allocated to each cell. This uniform channel distribution is efficient if the traffic distribution of the system is also uniform. In that case, the overall blocking probability of the mobile system is the same as the call blocking probability in a cell. Because traffic in cellular systems can be nonuniform with temporal and spatial fluctuations, a uniform allocation of channels to cells may result in high blocking in some cells, while others might have a sizeable number of spare channels. This could result in poor channel utilization. It is therefore appropriate to tailor the number of channels in a cell to match the load in it by nonuniform channel allocation [13, 14] or static borrowing [15, 16]. In nonuniform channel allocation the number of nominal channels allocated to each cell depends on the expected traffic profile in that cell. Thus, heavily loaded cells are assigned more channels than lightly loaded ones. In[13] an algorithm, namely nonuniform compact pattern allocation, is proposed for allocating channels to cells according to the traffic distribution in each of them. The proposed technique attempts to allocate channels to cells in such a way that the average blocking probability in the entire system is minimized. Let there be Ar cells and M channels in the system. The allocation of a channel to the set of co-channel cells forms a pattern which is referred to as the allocation pattern [13]. In addition, the compact allocation pattern of a channel is defined as the pattern with minimum average distance between cells. Given the traffic loads in each of the N cells and the possible compact pattern allocations for the M channels, the nonuniform compact pattern allocation algorithm attempts to find the compatible compact patterns that minimize the average blocking probability in the entire system as nominal channels are assigned one at a time. A similar technique for nonuniform channel allocation is also employed in the algorithms proposed in [14]. Simulation results in [13] show that the blocking probability using nonuniform compact pattern allocation is always lower than the blocking probability of uniform channel allocation. It is interesting to note that the reduction of blocking probability is almost uniformly 4 percent for the range of traffic shown in [13].2 Also for the same blocking probability, the


73

system can carry, on the average, 10 percent (maximum 22 percent) more traffic with the use of the nonuniform pattern allocation [13]. In the static borrowing schemes proposed in [15, 16], unused channels from lightly loaded cells are reassigned to heavily loaded ones at distances > the minimum reuse distance a. Although in static borrowing schemes channels are permanently assigned to cells, the number of nominal channels assigned in each cell may be reassigned periodically according to spatial inequities in the load. This can be done in a scheduled or predictive manner, with changes in traffic known in advance or based on measurements, respectively.

Channel Borrowing Schemes

In a channel borrowing scheme, an acceptor cell that has used all its nominal channels can borrow free channels from its neighboring cells (donors) to accommodate new calls. A channel can be borrowed by a cell if the borrowed channel does not interfere with existing calls. When a channel is borrowed, several other cells are prohibited from using it. This is called channel locking. The number of such cells depends on the cell layout and the type of initial allocation of channels to cells. For example, for a hexagonal planar layout with reuse distance of one cell (σ= 3), a borrowed channel is locked in three additional neighboring cells as is shown in Figure 4.2 while for a one-dimensional layout or a hexagonal planar grid layout with two cell reuse distance it is locked in two additional neighboring cells.

Fig. 4.2. Channel Locking

In contrast to Static Borrowing, channel borrowing strategies deal with short term allocation of borrowed channels to cells, and once a call is completed the borrowed channel is returned to its nominal cell. The proposed

74


Channel Borrowing schemes differ in the way a free channel is selected from a donor cell to be borrowed by an acceptor cell. The channel borrowing schemes can be divided into simple and hybrid. In the Simple Channel Borrowing schemes any nominal channel in a cell can be borrowed by a neighboring cell for temporary use. In the Hybrid Channel Borrowing strategies, the set of channels assigned to each cell is divided into two subsets A (standard or local channels) and B (non-standard or borrowable channels). The subset A is for use only in the nominally assigned cell, while the subset B is allowed to be lent to neighboring cells. Table 1 summarizes the Channel Borrowing schemes proposed in the literature. The interested reader can find an exhaustive account on channel Borrowing Schemes both simple and hybrid in [1] 4.1.2.5 Dynamic Channel Allocation

Due to short term temporal and spatial variations of traffic in cellular systems, FCA schemes are not able to attain high channel efficiency. To overcome this, Dynamic Channel Allocation (DCA) schemes have been studied during the past twenty years. In contrast to FCA, there is no fixed relationship between channels and cells in DCA. All channels are kept in a central pool and are assigned dynamically to radio cells as new calls arrive in the system [17, 18]. After a call is completed, its channel is returned to the central pool. In DCA, a channel is eligible for use in any cell provided that signal interference constraints are satisfied. Since, in general, more than one channel might be available in the pool. 1) Centralized DCA Schemes In the centralized DCA schemes, a channel from the central pool is assigned to a call for temporary use by a centralized controller. The difference between these schemes is the specific cost function used for selecting one of the candidate channels for assignment. First Available (FA): Among the DCA schemes the simplest one is the First Available (FA) strategy. In FA the first available channel within the reuse distance a encountered during a channel search is assigned to the call. The FA strategy minimizes the system computational time and as is shown by simulation in [12], for a linear cellular mobile system, it provides an increase of 20 % in the total handled traffic compared to FCA for low and moderate traffic loads. Locally Optimized Dynamic Assignment (LODA): In the Locally Optimized Dynamic Assignment (LODA) strategy [9, 20] the selected cost


75

function is based on the future blocking probability in the vicinity of the cell where a call is initiated. Channel Reuse Optimization Schemes

The objective of any mobile system is to maximize the efficiency of the system. Maximum efficiency is equivalent to maximum utilization of every channel in the system. It is obvious that the shorter the channel reuse distance, the greater the channel reuse over the whole service area. The cost functions selected in the following schemes attempt to maximize the efficiency of the system by optimizing the reuse of a channel in the system area. The schemes are differentiated on the Selection with Maximum Usage on the Reuse Ring, Mean Square (MSQ) , Nearest Neighbor (NN), Nearest Neighbor plus one (N N+1). (RING): In the Selection with Maximum Usage on the Reuse Ring (RING) strategy [12], a candidate channel is selected which is in use in the most cells in the co-channel set. If more than one channel has this maximum usage, an arbitrary selection among such channel is made to serve the call. If none is available, then the selection is made based on the FA scheme. Mean Square (MSQ), Nearest Neighbor (NN), Nearest Neighbor plus one (NN + 1): The Mean Square (MSQ) scheme selects the available channel that minimizes the mean square of the distance among the cells using the same channel. The Nearest Neighbor (NN) strategy selects the available channel occupied in the nearest cell in distance > a while the Nearest Neighbor plus one (NN + 1) scheme selects an eligible channel occupied in the nearest cell within distance > a + 1 or in distance a if an available channel is not found in distance a + 1 [12].

Performance Comparison Compared to FCA schemes, DCA schemes do not carry as much traffic at high blocking rates because they are not able to maximize the channel reuse as they serve the randomly offered call attempts. In order to improve the performance of DCA schemes at large traffic conditions, channel reassignment techniques have been suggested [15], [12], [21]. The basic goal of channel reassignment is to switch calls already in process, whenever possible, from channels that these calls are using, to other channels, with the objective of keeping the distance between cells using the same channel simultaneously to a minimum. Thus, the channel reuse is more concentrated and more traffic can be carried per channel at a given blocking rate.

76


Distributed DCA Schemes

Micro-cellular systems have shown a great potential for capacity improvement in high density personal communication networks [58, 17, 24]. However, propagation characteristics will be less predictable and network control requirements will be more intense than in the present systems. Several simulation and analysis results have shown that centralized DCA schemes can produce near optimum channel allocation, but at the expense of a high centralization overhead [8, 22, 23, 24, 25]. Distributed schemes are therefore more attractive for implementation in the micro-cellular systems, due to the simplicity of the assignment algorithm in each base station. The proposed distributed DCA schemes use either local information about the current available channels in the cell’s vicinity (Cell Based) schemes or signal strength measurements [26, 27, 6]. In the Cell Based schemes a channel is allocated to a call by the base station where the call is initiated. The difference with the centralized approach is that each base station keeps information about the current available channels in its vicinity. The channel pattern information is updated by exchange of status information between base stations. The Cell Based scheme provides near optimum channel allocation at the expense of excessive exchange of status information between base stations, especially under heavy traffic loads. 1) Signal Strength Measurement Based Distributed DCA Schemes A large body of research has been published on the performance analysis of channel allocation schemes, both FCA and DCA, [28,29,10,30,31], in which knowledge about the mobiles locations is not taken into account. In all of these schemes, channels are allocated to cells based on Sequential Channel Search (SCS): The simplest scheme among the interference adaptation DCA schemes is the Sequential Channel Search (SCS) strategy [26] where all mobile/ base station pairs examine channels in the same order and choose the first available with acceptable CIR. It is expected that SCS will support a volume of traffic by sub-optimal channel packing at the expense of causing many interruptions. A complete account of these schemes is given in [1]. In general, for the same blocking rate, DCA has a lower forced call termination rate than FCA. In FCA a call must be handed-off into another channel at every hand-off because the same channel is not available in adjacent cells. In DCA, the same channel can be assigned in the new cell if cochannel interference does not occur. In micro-cellular systems, mobiles cross cell boundaries frequently and the traffic of each cell varies drastically. Thus, a large amount of channel assignment control is required which results in frequent invocation of network control functions. Application of DCA schemes


77

in these system will be advantageous in solving the above problems due to flexibility in their channel assignment.[1] Hybrid Channel Allocation

The Hybrid Channel Assignment (HCA) schemes are a mixture of the FCA and DCA techniques. In HCA, the total number of channels available for service is divided into fixed and dynamic sets. The fixed set contains a number of nominal channels that are assigned to cells as in the FCA schemes and in all cases are to be preferred for use in their respective cells. The second set of channels is shared by all users in the system to increase flexibility. When a call requires service from a cell and all of its nominal channels are busy, then a channel from the dynamic set is assigned to the call. The channel assignment procedure from the dynamic set follows any of the DCA strategies described in the previous section. For example in the studies presented in [22] and [33], the FA and RING strategies are used respectively for the dynamic channel assignment. Variations of the main HCA schemes include HCA with channel reordering [14], and HCA schemes where calls that cannot find an available channel are queued instead of being blocked [78]. The call blocking probability for an HCA scheme is defined as the probability that a call arriving to a cell finds both the fixed and dynamic channels busy. Flexible Channel Allocation

In the Flexible Channel Allocation Schemes (FICA), the set of available channels is divided into fixed and flexible sets. Each cell is assigned a set of fixed channels that typically suffices under a light traffic load. The flexible channels are assigned to those cells whose channels have become inadequate under increasing traffic loads. The assignment of these emergency channels among the cells is done either in a scheduled or predictive manner [34]. In the literature proposed FICA techniques differ according to the time at which and the basis on which additional channels are assigned. More information is given in [1]. Fixed and Dynamic Channel Allocation

Fixed and Dynamic Channel assignment is a combination of FCA and DCA which tries to realize the lower of each technique’s blocking rate depending on traffic intensity. In low traffic intensity, the DCA scheme is used while in heavy traffic situations the FCA strategy is used. The transition from one strategy to the other should be done gradually because a sudden

78


transition will cause a lot of blocking. In [35], the authors developed an optimization model involving a single channel , a donor group, and an acceptor group of cells. An explicit formula is derived for the value of the load below dynamic assignment of the channel from the donor group to the acceptor group to minimize the overall blocking probability. This study analytically validates the belief that a strategy for dynamic channel assignment should be sensitive to the load of the system and yields an important insight in that the dynamic channel assignment should be disallowed in certain situations even if channels are free. The fixed and dynamic strategies allow assignment of channels in a dynamic fashion only if a minimum number of channels are free. This number depends on the value of the measured load. As the load increases, the minimum number of channels decreases and eventually under heavy loads the scheme starts to resemble the fixed allocation scheme [35]. Handling Hand-offs All the allocation schemes presented in the previous sections did not take into account the effect of hand-offs in the performance of the system. Handoff is defined as the change of radio channel used by a wireless terminal. The new radio channel can be with the same base station (intra-cell hanfoff) or with a new base station (inter-cell handoff). Adaptive Channel Allocation Reuse Partitioning Schemes

Several researchers have investigated Adaptive Channel Allocation (ACA) RUP schemes in an attempt to avoid the drawbacks of the Fixed RUP schemes [1]. With ACA Reuse Partitioning, any channel in the system can be used by any base station, as long as the required carrier to interference ratio (CIR) is maintained. It should be noted that reducing the CIR margin in each channel leads to an improvement in the traffic handling capacity. Based on this fact, a number of approaches such as flexible reuse schemes [1] and self organizing schemes [1] have been proposed. In [1], the autonomous reuse partitioning (ARP) was proposed, which assigns to a call the first channel found to exceed a CIR threshold in an ordered sequential channel search for each cell. The ARP technique was further improved in another scheme called Flexible Reuse, in which the channel with the minimum CIR margin is assigned [68]. Another scheme based on the ARP concept, called the Distributed Control Channel Allocation (DCCA) scheme, was proposed in [1]. In [1], the all channel concentric allocation (ACCA) which is an improved distributed version of the RUP scheme was proposed. Another scheme, namely the self organized reuse partitioning (SORP) which is based on signal power measurements at each station was proposed in [26]. In


79

[1] the channel assignment under the RUP concept was formulated as an optimization problem that maximizes the number of served calls. In the following, we provide a detailed description and discussion of the abovementioned reuse partitioning schemes. Overlaying Macrocellular Scheme

In micro-cellular systems, frequent hand-offs are very common. A channel assignment scheme different form the schemes discussed thus far is the overlay scheme. Here, a cluster of micro-cells are grouped together and covered by a macro-cell [1]. In overlay schemes, the total wireless resource is divided between the macro cell and all the micro cells in its domain. In the case of congestion, ifthere are not enough micro-cell channels for handoff calls, then macro-cell channels can be used. Since the macro-cell base station covers a much larger area compared to a micro-cell, its transmitted power is higher than micro-cell base stations. In the past, different channel assignment schemes for overlay cellular systems based on FCA and DCA schemes have been studied. In [1], a micro-cellular cluster having contiguous highway micro-cells each with its own BS is considered. Overlaying the micro-cellular cluster is a macro-cell whose BS also fulfills the role of the mobile switching center (MSC) of the micro-cellular cluster.

80


4.2 Channel Assignment Optimization

4.2.1 Introduction Today, mobile communications systems including TETRA experience a rapid increase in the number of subscribers, which places extra demands on their capacity. This growth leads to a new network architecture where the cells are designed to be increasingly smaller. The most serious problem that arises in this architecture is the handoff issue, which occurs when a mobile user moves from one microcell to a neighbouring one [36]. The majority of existing terrestrial wireless communication systems are based on the cellular concept [37, 38]. The underlying network structure is composed of a fixed network with wireless last hops between Base Stations (BSs) and Mobile Terminals (MTs). The fixed communication network connects the base stations to controllers. Mobile Switching Centers (MSCs), or Switching and Management Infrastructure for the case of TETRA (SwMI), that manage the calls and track all mobile terminal activities in a cell [39,40]. In some systems, multiple base stations are used to serve the same area. Hence, a multi-layer cellular network is formed [41,42]. The problem of handoff issue becomes more serious, for Ultra High and High Speed Moving Terminals where the handoff rate increases and the probability that an ongoing call will be dropped due to the lack of a free traffic channel is high. In the international literature, a great effort has been spent in order to study the handoff process and to minimize the involved handoff blocking probability [43]. The handoff blocking probability is considered to be more important than the blocking probability of new calls because the call is already active and the QoS is more sensitive for the handoff calls.

4.2.2 Model Formulation By taking into account that C are the available channels in every microcell, C channels are shared both by new and handoff calls. The following assumptions, without affecting the results, are considered: The terminals are characterized as Low Speed Moving Terminals (LSMT), High Speed Moving Terminals (HSMT) or Ultra High Speed Moving Terminals (UHSMT) according to the speed they move. The speed (velocity) of a mobile is measured approximately by simply gathering the time spent in


81

a cell by a mobile. A more accurate estimation of the terminals’ speed is possible if the received Doppler frequency is known. There is a useful relationship between the branch switching rate of diversity receiver and its Doppler frequency, which permits the estimation of vehicle speed without significant hardware changes [45]. Also, homogenous traffic, same capacity and same mean holding time Th are considered in all microcells. The model will cover both one layer and three layer architectures. New and handoff calls of LSMT are generated in the area of microcell according to a Poisson point process, with mean rates of Λ LR , Λ LRh respectively, while new calls and handoff calls of HSMT are generated with mean rates of Λ LR , Λ LRh and new calls and handoff calls of UHSMT are UH generated with mean rates of ΛUH R , Λ Rh per cell. The relative mobilities are defined as:

aL =

Λ LRh for LSMT Λ LRh + Λ LR

(4.3)

aH =

Λ HRh for HSMT Λ HRh + Λ HR

(4.4)

ΛUH Rh for UHSMT ΛUH + ΛUH Rh R

(4.5)

aUH =

Also is defined the coefficient that represents the traffic load of Low Speed Moving Terminals toward to traffic load of all calls generated per cell :

kL =

Λ LR + Λ LRh UH Λ LR + Λ LRh + Λ HR + Λ HRh + ΛUH R + Λ Rh

(4.6)

The coefficient that represents the traffic load of High Speed Moving Terminals toward to traffic load of all calls generated per cell is:

Λ HR + Λ HRh kH = L UH Λ R + Λ LRh + Λ HR + Λ HRh + ΛUH R + Λ Rh

(4.7)

and lastly the coefficient that represents the traffic load of Ultra High Speed Moving Terminals toward to traffic load of all calls generated per cell is:

82


kUH =

UH ΛUH R + Λ Rh UH Λ LR + Λ LRh + Λ HR + Λ HRh + ΛUH R + Λ Rh

(4.8)

The offered load per cell is

Toff =

UH Λ LR + Λ LRh + Λ HR + Λ HRh + ΛUH R + Λ Rh

µΗ

(4.9)

where µH=1/TH and TH is the channel holding time. 4.2.3 One Layer Architecture using Erlang Model Let n be the number of microcells in the microcellular area. The total offered load in the system is:

Tofftot = n ⋅ Toff

(4.10)

and the total number of channels in the system is:

C = n ⋅ Cm

(4.11)

The steady state probabilities that j channels are busy in every microcell, can be derived from Figure 4.3. [44]

Fig. 4.3 State Transition diagram for microcells for the Existing One Layer Architecture using Erlang Model where: UH Λ = Λ LR + Λ LRh + Λ HR + Λ HRh + ΛUH R + Λ Rh

(4.12)


83

⎧ ( Λ L + Λ L + Λ H + Λ H + ΛUH + ΛUH ) j Rh R Rh R Rh ⎪ R P0 ⎪⎪ j !µΗ j Pj = ⎨ j − ( C −C ) UH Cm −Chm ⎪ ( Λ LR + Λ LRh + Λ HR + Λ HRh + ΛUH ( Λ LRh + ΛHRh + ΛUHRh ) m hm P R + Λ Rh ) ⎪ 0 (4.13) j !µΗ j ⎪⎩

for j = 1, 2,..., Cm − Chm for j = Cm − Chm + 1,...Cm where: ⎡ Cm −Chm ( Λ L + Λ L + Λ H + Λ H + ΛUH + ΛUH )k R Rh R Rh R Rh + P0 = ⎢ ∑ ⎢ k =0 k !µH k ⎣ Cm

∑

(Λ

L R

)

UH Cm −Chm Rh

+ ΛLRh + ΛHR + ΛHRh + ΛUH R +Λ

k !µH

k =Cm −Chm +1

(Λ

L Rh

(4.14)

)

UH k −( Cm −Chm ) Rh

+ ΛHRh + Λ

k

⎤ ⎥ ⎥ ⎦

−1

The blocking probability (PB) for a new call (either UHSMT, HSMT or LSMT) per microcell is the sum of probabilities that the state number (j) of the microcell is greater than (Cm-Chm). Hence:

PB =

Cm

∑

j = Cm − Chm

Pj

(4.15)

The blocking probability for a new call of UHSMT per microcell is: UHSMT B

P

ΛUH R = L PB H Λ R + Λ R + ΛUH R

(4.16)

The blocking probability for a new call of HSMT per microcell is: HSMT B

P

Λ HR = L PB Λ R + Λ HR + ΛUH R

(4.17)

and the blocking probability for a new call of LSMT is:

PBLSMT =

Λ LR PB Λ LR + Λ HR + ΛUH R

(4.18)

84


The probability of handoff attempt failure Pfh is the probability that the state number of the microcell is equal to Cm. Thus:

Pfh = PCm

(4.19)

The handoff blocking probability of UHSMT is:

PfhHSMT =

ΛUH Rh Pfh Λ LRh + Λ HRh + ΛUH Rh

(4.20)

The handoff blocking probability of HSMT is:

PfhHSMT =

Λ HRh Pfh Λ LRh + Λ HRh + ΛUH Rh

(4.21)

an the handoff blocking probability of LSMT is:

PfhLSMT =

Λ LRh Pfh Λ LRh + Λ HRh + ΛUH Rh

(4.22)

In this model, the ratio of guard channels toward the total available channels in every microcell is fixed. 4.2.4 Channel Assignment Scheme based on a Three Layer Architecture A new channel assignment model is proposed in order to determine the optimized number of channels that should be assigned to satellite cell, to macrocells and to microcells. The purpose of this optimized determination is to decrease the Quality of Service (QoS) of both HSMT and UHSMT with the smallest possible effect on the QoS of LSMT. A multi-layer architecture is introduced in order to dedicate different layers to different types of subscribers according to their speed in the same geographical area, [46-52] as it is shown in Figure 2. The implementation of the different layers doesn’t require any special hardware setting but only new radio parameters in the existing software. This approximation, introduces three-layer architecture, the microcellular layer, the macrocellular layer and the satellite layer. In addition, the microcell layer services only new and handoff calls of LSMT, the macrocell layer services new and handoff calls of HSMT and lastly the satellite layer services new and handoff calls of UHSMT. Figure 4.3 shows the call assignment in the


85

different layers. Also, homogeneous traffic and same Th is considered in microcells, macrocells and satellite cells.

Satellite Cell

Macro cell 1

Micro cell 1 Micro cell 2

...

Macro cell 2

Micro cell n


...

Macro cell m

Micro cell n


...

Micro cell n

Fig. 4.4 Multi Layer Network

New and Handoff calls of UHSMT

New and Handoff calls of HSMT New and Handoff calls of LSMT

Satellite Cell

Macro cell 1


...

Micro cell n

Macro cell 2


...

Micro cell n

Macro cell m


...

Micro cell n

Fig. 4.5. Call Assignment in Different Layers

Let m be the number of macrocells that are under the satellite cell and consist the macrocellular layer and n the number of microcells that are under every macrocell. Let C be the total number of channels in the system. In the microcellular layer, priority is given to handoff attempts by assigning guard channels (Chm) exclusively for handoff calls of LSMT among the Cm channels in a cell. The remaining (Cm-Chm) channels are shared by both new and handoff calls of LSMT [44]. In the macrocellular layer, priority is given to handoff attempts by assigning guard channels (ChM) exclusively for handoff calls of HSMT among the CM channels in a cell. The remaining

86


(C M-ChM) channels are shared by both new and handoff calls of HSMT. Let CS be the channels assigned to satellite cell. Priority is given to handoff attempts by assigning guard channels (ChS) exclusively for handoff calls of UHSMT among the CS channels in the umbrella cell. The remaining (CSChS) channels are shared by both new and handoff calls of UHSMT [44]. Hence:

C = C S + m ⋅ C M + m ⋅ (n ⋅ C m )

(4.23)

The relative ratios for the guard channels toward the available channels for the satellite cell is:

gc S =

C hS CS

(4.24)

gc M =

C hM CM

(4.25)

gc m =

C hm Cm

(4.26)

for every macrocell is:

and for every microcell is:

The mean rate of generation of new and handoff calls of UHSMT is Λ and ΛUH Rh per cell, so the mean rate generated in the macrocell is UH R

n ⋅ m ⋅ ΛHR and n ⋅ m ⋅ ΛHRh respectively. The mean rate of generation of new and handoff calls of HSMT is ΛHR and ΛHRh respectively per cell, so the mean rate generated in the macrocell is m ⋅ ΛHR and m ⋅ ΛHRh respectively. Lastly, the mean rate of generation of new and handoff calls of LSMT is ΛLR and ΛLRh per cell. The proposed channel assignment scheme, assigns the ratios both gCS gCM, gCm and Cm/C, CM/C, CS/C according to αL, αH, αUH, kL, kH, kUH and Tofftot , contributing to the improvement of the QoS of both UHSMT and HSMT, with the smallest possible effect on the QoS of LSMT. The steady state probabilities that j channels are busy in a microcell can be derived from figure 4.6 [37],[44]


87

Fig. 4.6 State Transition diagram for every microcell in proposed channel assignment technique

⎧(ΛL + ΛL ) j Rh ⎪ R P0m j ⎪⎪ j ! µ H Pjm = ⎨ Cm − Chm j − C −C ⎪ ( Λ LR + Λ LRh ) Λ LRh ( m hm ) m P0 ⎪ j !µH j ⎪⎩

(4.27)

for j = 1,2,..., C m − C hm for j = C m − C hm + 1,..., C m where Cm −Chm L k −( Cm −Chm ) ⎤ ⎡Cm −Chm ( Λ L + Λ L )k Cm ΛLR + ΛLRh ) ΛRh ( R Rh ⎥ ⎢ P = ∑ + ∑ ⎥ ⎢ k =0 k !µH k k !µH k k =Cm −Chm +1 ⎦ ⎣ m 0

−1

(4.28)

The blocking probability for a new call of LSMT per microcell is the sum of probabilities that the state number of the microcell is greater than CmChm. Hence:

PBm =

Cm

∑

j =Cm − Chm

Pjm

(4.29)

The probability of handoff attempt failure Pfhm is the probability that the state number of the microcell is equal to Cm. Thus: Pfhm = PCm

(4.30)

88


For the macrocell, the steady state probabilities that j channels are busy can be derived from figure 4.7 [37],[44]:

Fig. 4.7 State Transition diagram for every macrocell in proposed channel assignment technique

⎧ ( nΛ H + nΛ H ) j R Rh ⎪ P0M j ⎪⎪ j !µH PjM = ⎨ CM − ChM j − ( CM − ChM ) ⎪ ( nΛ HR + nΛ HRh ) nΛ HRh ) ( P0M ⎪ j j !µH ⎪⎩

(4.31)

for j = 1,2,..., C M − C hM for j = C M − C hM + 1,..., C M where

⎡CM −ChM ( nΛ H + nΛ H )k R Rh + P =⎢ ∑ k ⎢ k =0 k ! µH ⎣ M 0

(4.32) CM

∑

k =CM −ChM +1

( nΛ

H R

+ nΛHRh )

CM −ChM

( nΛ )

k !µH k

H Rh

k −( CM −ChM )

⎤ ⎥ ⎥ ⎦

−1

The blocking probability for a new call of HSMT in the umbrella cell is the sum of probabilities that the state number of that cell is greater than CM-ChM. Hence:


PBM =

89

CM

∑

j = CM − ChM

PjM

(4.33)

The probability that a handoff call will be blocked in the umbrella cell is PfhM and is the probability that state number of the cell is equal to Cu. Thus: M PfhM = PCM

(4.34)

For the satellite cell, the steady state probabilities that j channels are busy can be derived from figure 4.8 [37],[44]:

Fig. 4.8 State Transition diagram for satellite cell in proposed channel assignment technique

⎧ ( nmΛUH + nmΛUH ) j R Rh ⎪ P0S j ⎪⎪ j !µH PjS = ⎨ j −( CS −ChS ) UH CS − ChS ⎪ ( nmΛUH nmΛUH ( R + nmΛ Rh ) Rh ) P0S ⎪ j j !µH ⎪⎩ for j = 1, 2,..., CS − ChS for j = CS − ChS + 1,..., CS where

(4.35)

90


(

⎡CS −ChS nmΛUH + nmΛUH R Rh P =⎢ ∑ k k!µ H ⎢⎣ k =0 u 0

CS

∑

( nmΛ

UH R

)

UH CS −ChS Rh

+ nmΛ

k

( nmΛ )

k !µH k

k =CS −ChS +1

)

+

UH k −( CS −ChS ) Rh

⎤ ⎥ ⎥ ⎦

−1

(4.36)

The blocking probability for a new call of HSMT in the umbrella cell is the sum of probabilities that the state number of that cell is greater than CSChS. Hence:

PBS =

CS

∑

j =CS − ChS

PjS

(4.37)

The probability that a handoff call will be blocked in the umbrella cell is PfhS and is the probability that state number of the cell is equal to CS. Thus:

PfhS = PCSS

(4.38)

The measure of system performance (Quality of Service) is a cost function [48-66] which uses system system’s data as the average new call origination rate and the average handoff attempt rate per cell. This cost function can be expressed as:

CF =

Λ R ⋅ PB + Λ Rh ⋅ Pfh Λ R + Λ Rh

(4.39)

Therefore, the QoS for calls especially for both UHSMT and HSMT must be guaranteed while allowing high utilization of channels. The objective of the proposed technique based on the three-layer architecture is to guarantee the required QoS of both UHSMT and HSMT. 4.2.5 Comparison of One layer with Three Layer Architecture

A new channel assignment model thus proposed in order to determine the optimized number of channels that should be assigned microcells, to macrocells and to satellite cell. Applying this model to a cellular system and for that matter to TETRA, a better performance both of UHSMT and


91

HSMT is achieved. In this model using the three-layer architecture, the satellite cell and the macrocells have been introduced to serve calls of UHSMT and HSMT respectively. Moreover, according to the obtained results, the cost functions of UHSMT and HSMT have been optimized having a minimum effect on the cost function of LSMT. The comparison of the existing model based on an One Layer Architecture and using the Erlang Model (OLA-EM) with the Proposed Channel Assignment Model Based on a Three Layer Architecture (PCAM-TLA) has been done for different values of CS/C, Cm/C, CM/C, kL, kH, kUH, gCS, gCM and gCm. The figures below present the behaviour of the techniques of previous sections for the optimized values of above variables. In the performed simulation the following parameters are considered without affecting the generality of the model: The number of satellite cell is considered to be one, n=3, m=3, TH=80s, CS=240, gCu=0.10, gCS=0.10, gCM=0.10, αL=0.6, αH=0.4, αUH=0.45, kUH=0.25, kL=0.2, kH=0.35,. Using T tot

these values and 0≤ off ≤300 erlangs, the Λ LR , Λ LRh Λ HR , Λ HRh are calculated. In all figures curve (i) represents the performance of a typical cellular system using Erlang Model for Cs=240. In this case, there is only one layer and all the involved calls are served by microcells. Curves (ii), (iii) and (iv) show the performance of Proposed Channel Assignment Model based on a three-layer architecture, for CS=108, CM=36, Cm=6, (ii) CS=90, CM=42, Cm=6 and (iii) CS=72, CM=48, Cm=6 respectively Figure 4.9 presents the QoS of LMT for Low Traffic Conditions. Figure 4.10 presents the QoS of HMT and figure 4.11presents the QoS of UHSMT for Low Traffic Conditions. Figure 4.12 presents the QoS of HSMT and UHSMT for Low Traffic Conditions

92


10

Cost Function

10

10

10

10

10

0

-20

-40

-60

-80

-100

0

10 20 30 40 Total Offered Traffic Load in System (Erlangs)

50

Fig. 4.9. Cost Function for LSMT (Low Traffic Conditions)

10

Cost Function

10

10

10

10

10

10

-20

-40

-60

-80

-100

-120

-140

0


Fig. 4.10. Cost Function for HSMT (Low Traffic Conditions)

50


10

Cost Function

10

10

10

10

10

10

0

-50

-100

-150

-200

-250

-300

0


50

Fig 4.11. Cost Function for UHSMT (Low Traffic Conditions)

10

10

Cost Function

10

10

10

10

10

-20

-40

-60

-80

-100

-120

-140

0


50

Fig. 4.12. Cost Function for UHSMT and HSMT (Low Traffic Conditions)

93

94


Figure 4.13 presents the QoS of LMT for Medium Traffic Conditions. Figure 4.14 presents the QoS of HMT and Figure 4.15 presents the QoS of UHSMT for Medium Traffic Conditions. Figure 4.16 presents the QoS of HSMT and UHSMT for Medium Traffic Conditions. 10

Cost Function

10

10

10

10

0

-10

-20

-30

-40

50

100 150 Total Offered Traffic Load in System (Erlangs)

Fig. 4.13. Cost Function for LSMT (Medium Traffic Conditions)

200


10

Cost Function

10

10

10

10

10

-20

-30

-40

-50

-60

-70

50


200

Fig. 4.14. Cost Function for HSMT (Medium Traffic Conditions)

10

Cost Function

10

10

10

10

10

-20

-30

-40

-50

-60

-70

50


Fig. 4.15. Cost Function for UHSMT (Medium Traffic Conditions)

200

95

96


10

Cost Function

10

10

10

10

10

-20

-40

-60

-80

-100

-120

50


200

Fig. 4.16. Cost Function for UHSMT and HSMT (Medium Traffic Conditions)

Figure 4.17 presents the QoS of LMT for Low Traffic Conditions. Figure 4.18 presents the QoS of HMT and Figure 4.19 presents the QoS of UHSMT for Low Traffic Conditions. Figure 4.20 presents the QoS of HSMT and UHSMT for Low Traffic Conditions.


10

Cost Function

10

10

10

10

10

0

-5

-10

-15

-20

-25

200


300

Fig. 4.17. Cost Function for LSMT (High Traffic Conditions)

10

Cost Function

10

10

10

10

10

-15

-20

-25

-30

-35

-40

200


Fig. 4.18. Cost Function for HSMT (High Traffic Conditions)

300

97

98


10

Cost Function

10

10

10

10

10

-10

-20

-30

-40

-50

-60

200


300

Fig 4.19. Cost Function for UHSMT (High Traffic Conditions)

Cost Function

10

10

10

10

-15

-20

-25

-30

200


300

Fig. 4.20. Cost Function for UHSMT and HSMT (High Traffic Conditions)

Curves of all figures show an improvement in the Cost Function of UHSMT and HSMT as a result of using the proposed technique, as well as


99

adjusting the ratios CS/C, Cm/C, CM/C according to the Tofftot , αL,αH,αUH, kUH, kL and kH. This improvement depends both on the number of channels that are assigned to every layer. All figures show that the above ratios optimize the behaviour of the UHSMT and HSMT with the minimum bad effect on the QoS of of LSMT. Figure 4.21 presents the Cost Function that relates to the Low Speed Moving Terminals. Connaturally, figure 4.22 presents the Cost Function for High Speed Moving Terminals, figure 4.23 presents the Cost Function for Ultra High Speed Moving Terminals and lastly figure 4.24 presents the Cost Function for High and Ultra High Speed Moving Terminals. 0

10

-20

Cost Function

10

-40

10

-60

10

-80

10

-100

10

0

50


Fig. 4.21. Cost Function for Low Speed Moving Terminals

300

100 4 Channel Assignment and Multiple Access in Trunking Radio Systems [1]

0

10

-20

10

-40

Cost Function

10

-60

10

-80

10

-100

10

-120

10

-140

10

0

50


300

Fig. 4.22. Cost Function for High Speed Moving Terminals 0

10

-50

10

-100

Cost Function

10

-150

10

-200

10

-250

10

-300

10

0

50


Fig. 4.23. Cost Function for Ultra High Speed Moving Terminals

300


0

10

-20

10

-40

Cost Function

10

-60

10

-80

10

-100

10

-120

10

-140

10

0

50


300

Fig. 4.24. Cost Function for Ultra High and High Speed Moving Terminals

101


4.3 Multiple Access Techniques

4.3.1 CDMA Techniques in TETRA systems Introduction

A TETRA mobile communication network is a multi-user system, in which a large number of users share a common physical resource to transmit and receive information. Multiple access capability is one of the fundamental components. The spectral spreading of the transmitted signal gives the feasibility of multiple access to CDMA systems. Figure 4.25 shows three different and commonly used multiple access technologies: TDMA, FDMA and CDMA.

Fig. 4.25. Multiple Access Technologies

In FDMA, (Frequency Division Multiple Access), signals for different users are transmitted in different channels each with a different modulating frequency; in TDMA, (Time Division Multiple Access), signals for different users are transmitted in different time slots. With these two technologies, the maximum number of users who can share the physical channels simultaneously is fixed. However, in CDMA, signals for different users are transmitted in the same frequency band at the same time. Each user’s signal acts as interference to other user’s signals and hence the capacity of the CDMA system is related closely to the interference level: there is no fixed maximum number, so the term soft capacity is used. Figure 4.26 shows an example of how 3 users can have simultaneous access in a CDMA system.


103

At the receiver, user 2 de-spreads its information signal back to the narrow band signal, but nobody else’s. This is because that the cross-correlations between the code of the desired user and the codes of other users are small: coherent detection will only put the power of the desired signal and a small part of the signal from other users into the information bandwidth. The processing gain, together with the wideband nature of the process, gives benefits to CDMA systems, such as high spectral efficiency and soft capacity.

Fig. 4.26. Principle of spread-spectrum multiple access

However, all these benefits require the use of tight power control and soft handover to avoid one user’s signal cloaking the communication of others. Compared with the conventional hard handover, soft handover has the advantages of smoother transmission and less ping-pong effects. As well as leading to continuity of the wireless services, it also brings macrodiversity gain to the system. However, soft handover has the disadvantages of complexity and extra resource consumption. Therefore, optimisation is crucial for guaranteeing the performance of soft handover. In soft handover, the mobile station starts communications with a new base station without disconnecting from the old base station. The voice quality is improved due to the diversity provided by the extra channel path at the cell edge if compared with hard handover. From the TETRA network operator point of view, however, more resources are required to support this procedure. Until now, several algorithms have been proposed aiming at maximising the macrodiversity gain and minimising the handover failure rate and extensive research has been conducted on the optimisation of the parameters for these soft handover algorithms. The objective of this chapter is to develop an analytical model to study the performance of soft handover algorithms which resemble the type described in IS-95, but the details may vary. The analysis yields performance


measures, such as the number of active set updates, and the number of base stations occupied by one mobile and outage probability of the received signal. The analysis methodology is an extension of the methodology for hard handovers, as presented in bibliography [67] and [68]. Besides the algorithmic differences between hard and soft handovers, a significant difference exits between the performance metrics emphasized. In [67] and [68], the tradeoff between handover delay and “ping ponging” was emphasized. For soft handover, there is an important tradeoff between improved quality due to diversity and increased resource utilization during soft handover. References providing additional information on soft handover include [69]-[71], [73] and [74].These papers provide a more system-wide perspective, while we focus more on parameter selection for algorithm design. In this chapter, an analytical overview of CDMA technology is initially presented as well as an overview of the handover procedure in CDMA based systems. Subsequently the different types of Handover in TETRA systems and the objectives of Handover with emphasis in soft handover procedure, features and metrics are presented. Finally the analysis of the algorithms, numerical results based on the analysis presented and comparison of these data and the soft handover performance in different scenarios can be found in the end of the chapter. Principles of spectrum spreading (CDMA)

Digital communications systems are designed to maximise capacity utilisation. From Shannon’s channel capacity principle expressed as (4.40), it is obvious that the channel capacity can be increased by increasing the channel bandwidth.

(

C = B ⋅ log 2 1 + S

N

)

(4.40)

Where B is the bandwidth (Hz), C is the channel capacity (bit/s), S is the signal power and N is the noise power. Thus, for a particular S/N ratio (Signal to Noise Ratio: SNR), the capacity is increased if the bandwidth used to transfer information is increased. CDMA is a technology that spreads the original signal to a wideband signal before transmission. CDMA is often called as Spread-Spectrum Multiple Access (SSMA). The ratio of transmitted bandwidth to information bandwidth is called the processing gain Gp (also called spreading factor).


GP =

Bt BI

OR G P

=

B R

105

(4.41)

Where Bt is the transmission bandwidth, Bi is the bandwidth of the information bearing signal, B is the RF bandwidth and R is the information rate. Relating the S/N ratio to the Eb/I0 ratio, where Eb is the energy per bit, and I0 is the noise power spectral density, leading to:

S Eb × R Eb 1 = = × N I O × B I O GP

(4.42)

Therefore, for certain Eb/I0 requirement, the higher the processing gain, the lower the S/N ratio required. In the first CDMA system, IS-95, the transmission bandwidth is 1.25 MHz. In CDMA, each user is assigned a unique code sequence (spreading code) that is used to spread the information signal to a wideband signal before being transmitted. The receiver knows the code sequence for that user, and can hence decode it and recover the original data. Spreading and de-spreading

Spreading and de-spreading are the most basic operations in DS-CDMA systems, shown as Figure 3. User data here is assumed to be a BPSKmodulated bit sequence of rate R. The spread operation is the multiplication of each user data bit with a sequence of n code bits, called chips. Here, n=8, and hence the spreading factor is 8. This is also assumed for the BPSK spreading modulation. The resulting spread data is at a rate of 8×R and has the same random (pseudo-noise-like) appearance as the spreading code. The increase of data rate by a factor 8 corresponds to a widening (by a factor 8) of the occupied spectrum of the spread user data signal. This wideband signal would then be transmitted across a radio channel to the receiving end.


Fig. 4.27. Spreading and de-spreading

During de-spreading, the spread user data/chip sequence is multiplied bit by bit with the same 8 code chips as used during the spreading process. As shown, the original user’s data is recovered perfectly. Types of handover in CDMA systems

There are four different types of handovers in WCDMA mobile networks. They are: • Intra-system HO: Intra-system HO occurs within one system. It can be further divided into Intrafrequency HO and Inter-frequency HO. Intrafrequency occurs between cells belonging to the same WCDMA carrier, while Inter-frequency occurs between cells operate on different WCDMA carriers. • Inter-system HO: Inter-system HO takes places between cells belonging to two different Radio Access Technologies (RAT) or different Radio Access Modes (RAM). The most frequent case for the first type is expected between WCDMA and GSM/EDGE systems. Handover between two different CDMA systems also belongs to this type. An example of inter-RAM HO is between UTRA FDD and UTRA TDD modes. • Hard Handover (HHO): HHO is a category of HO procedures in which all the old radio links of a mobile are released before the new radio links are established. For real-time bearers it means a short disconnection of the bearer; for non-real-time bearers HHO is lossless. Hard handover can take place as intra or inter-frequency handover. • Soft Handover (SHO) and Softer HO: During soft handover, a mobile simultaneously communicates with two (2-way SHO) or more cells


107

belonging to different BSs of the same RNC (intra-RNC) or different RNCs (inter-RNC). In the downlink (DL), the mobile receives both signals for maximal ratio combining: in the uplink (UL), the mobile code channel is detected by both BSs (2-way SHO), and is routed to the RNC for selection combining. Two active power control loops participate in soft handover: one for each BS. In the softer handover situation, a mobile is controlled by at least two sectors under one BS, the RNC is not involved and there is only one active power control loop. SHO and softer HO are only possible within one carrier frequency and therefore, they are intra-frequency handover processes. Figure 4.28 shows some scenarios of different types of HO.

Fig. 4.28. Scenarios of different types of handover

1) Objectives of handover

Handover can be initiated in three different ways: mobile initiated, network initiated and mobile assisted. • Mobile Initiated: the Mobile makes quality measurements, picks the best BS, and switches, with the network’s cooperation. This type of handover is generally triggered by the poor link quality measured by the mobile. • Network Initiated: the BS makes the measurements and reports to the RNC, which makes the decision whether to handover or not. Network initiated handover is executed for reasons other than radio link control, e.g. to control traffic distribution between cells. An example of this is the BS-controlled Traffic Reason Handover (TRHO). TRHO is a loadbased algorithm that changes the handover threshold for one or more outgoing adjacencies for a given source cell depending on its load. If the


load of the source cell exceeds a given level, and the load in a neighbouring cell is below another given level, then the source cell will shrink its coverage, handing over some traffic to the neighbouring cell. Therefore, the overall blocking rate can be reduced, leading to a greater utilization of the cell resource. • Mobile Assisted: here the network and the mobile both make measurements. The mobile reports the measurement results from nearby BSs and the network makes the decision of handing over or not. The objectives of handover can be summarised as follows: • Guaranteeing the continuity of wireless services when the mobile user moves across the cellular boundaries • Keep required QoS • Minimising interference level of the whole system by keeping the mobile linked to the strongest BS or BSs. • Roaming between different networks • Distributing load from hot spot areas (load balancing)

The triggers that can be used for the initiation of a handover process could be the link quality (UL or DL), the changing of service and the changing of speed, traffic reasons or O&M (Operation & Maintenance) intervention. 2) Handover measurements and procedures

The handover procedure can be divided into three phases: measurement, decision and execution phases as illustrated in Figure 4.29. In the handover measurement phase, the necessary information needed to make the handover decision is measured. Typical downlink measurements performed by the mobile are the Ec/I0 of the Common Pilot Channel (CPICH) of its serving cell and neighbouring cells. For certain types of handover, other measurements are needed as well. For example, in an asynchronous network like UTRA FDD (WCDMA), the relative timing information between the cells needs to be measured in order to adjust the transmission timing in soft handover to allow coherent combining in the Rake receiver. Otherwise, the transmissions from the different BSs would be difficult to combine and especially the power control operation in soft handover would suffer additional delay.


109

Fig. 4.29. Handover procedures

In the handover decision phase, the measurement results are compared against the predefined thresholds and then it is decided whether to initiate the handover or not. Different handover algorithms have different trigger conditions. In the execution phase, the handover process is completed and the relative parameters are changed according to the different types of handover. For example, in the execution phase of the soft handover, the mobile enters or leaves the soft handover state, a new BS is added or released, the active set is updated and the power of each channel involved in soft handover is adjusted. Soft Handover (SHO)

ι) Principles of soft handover

Soft handover is different from the traditional hard handover process. With hard handover, a definite decision is made on whether to handover or not and the mobile only communicates with one BS at a time. With soft handover, a conditional decision is made on whether to handover or not. Depending on the changes in pilot signal strength from the two or more BSs involved, a hard decision will eventually be made to communicate with only one. This normally happens after it is clear that the signal coming from one BS is considerably stronger than those come from the others. In the interim period of soft handover, the mobile communicates simultaneously


with all the BSs in the active set, which is the list of cells that are presently having connections with the mobile. Planning of soft handover overhead is one of the fundamental components of the radio network planning and optimisation. In this section, the basic principles of soft handover are presented. The difference between hard and soft handover is like the difference between a swimming relay and track-and-field relay events. Hard handover happens on a time point; while, soft handover lasts for a period of time. Figure 6 shows the basic process of hard and soft handover (2-way case). Moreover, a CDMA TETRA system, soft handover has two opposing effects on the downlink: macrodiversity and extra resource consumption. Macrodiversity improves the link level performance, but for system level performance there is a trade-off between these two effects.[76-85] To maximise the downlink capacity, there is an optimum soft handover overhead. This optimum overhead is sensitive to the cell selection scheme, power control conditions and variety of the radio parameters. The size of the active set is not governed by “the bigger the better”. Considering the complexity and the increased signalling, that comes with adding an extra BS, when implementing soft handover, the size of the active set should be kept as two. The optimisation of soft handover shows that the capacity gain is related closely to the handover threshold. However, there is a tradeoff between the system capacity and the signalling load. By optimising the soft handover procedure, possible downlink bottleneck situation would be mitigated in future mobile networks. The simulation analysis presented in the last Sections of this chapter enables us to provide insightful results and it helps in interpreting more complex simulations needed for total system characterization. The introduction of the timer in soft-handover algorithm is effective in the sense that it reduces the system overhead heavily with only a slight increasing in resource usage. Possible extensions of the modelling include other IS-95 parameters, more than two base stations, etc.., with attending study of algorithmic and parametric robustness. Assuming there is a mobile terminal inside the car moving from cell 1 to cell 2, BS1 is the mobile’s original serving BS. While moving, the mobile continuously measures the pilot signal strength received from the nearby BSs. With hard handover shown as (a) in Figure 4.30, the trigger of the handover can be simply described as:


111

Where (pilot_ Ec/I0)1 and (pilot_ Ec/I0)2 are the received pilot Ec/I0 from BS1 and BS2 respectively; D is the hysteresis margin.

Fig. 4.30. Comparison between hard and soft handover

The reason for introducing the hysteresis margin in the hard handover algorithm is to avoid a “ping-pong effect”, the phenomenon that when a mobile moves in and out of cell’s boundary, frequent hard handover occurs. Apart from the mobility of the mobile, fading effects of the radio channel can also make the “ping-pong” effect more serious. By introducing the hysteresis margin, the “ping-pong” effect is mitigated because the mobile does not handover immediately to the better BS. The bigger margin, the less the “ping-pong” effect. However, a big margin means more delay. Moreover, the mobile causes extra interference to neighbouring cells due to the poor quality link during the delay. Therefore, to hard handover, the value of the hysteresis margin is fairly important. When hard handover occurs, the original traffic link with BS1 is dropped before the


setting up of the new link with BS2 so hard handover is a process of “break before make”. In the case of soft handover, shown as (b) in Figure 6, before (pilot_ Ec/I0)2 goes beyond (pilot_ Ec/I0)1, as long as the soft handover trigger condition is fulfilled, the mobile enters the soft handover state and a new link is set up. Before BS1 is dropped (handover dropping condition is fulfilled), the mobile communicates with both BS1 and BS2 simultaneously. Therefore, unlike hard handover, soft handover is a process of “make before break”. So far, several algorithms have been proposed to support soft handover and different criteria are used in different algorithms. The soft handover process is not the same in the different transmission directions. Figure 4.31 illustrates this. In the uplink, the mobile transmits the signals to the air through its omnidirectional antenna. The two BSs in the active set can receive the signals simultaneously because of the frequency reuse factor of one in CDMA systems. Then, the signals are passed forward to the RNC for selection combining. The better frame is selected and the other is discarded. Therefore, in the uplink, there is no extra channel needed to support soft handover.

Fig. 4.31. Principles of soft handover (2-way case)

In the downlink, the same signals are transmitted through both BSs and the mobile can coherently combine the signals from different BSs since it sees them as just additional multipath components. Normally maximum ratio combining strategy is used, which provides an additional benefit called macrodiversity. However, to support soft handover in the downlink, at least one extra downlink channel (2-way SHO) is needed. This extra


113

downlink channel acts to other users acts like additional interference in the air interface. Thus, to support soft handover in the downlink, more resource is required. As a result, in the downlink direction, the performance of the soft handover depends on the trade-off between the macrodiversity gain and the extra resource consumption.

ii) Algorithm of soft handover

The performance of soft handover is related closely to the algorithm. Figure 4.32 shows the IS-95A soft handover algorithm (also called basic cdmaOne algorithm) [EIA/TIA/IS-95A] [Qualcomm97] [L-SJSWK99].

Fig. 4.32. IS-95A soft handover algorithm

The active set is the list of cells that currently have connections to the mobile; the candidate set is the list of cells that are not presently used in the soft handover connection, but whose pilot Ec/I0 values are strong enough to be added to the active set; the neighbouring set (monitored set) is the list of cells that the mobile continuously measures, but whose pilot


Ec/I0 values are not strong enough to be added to the active set. In IS-95A, the handover threshold is a fixed value of received pilot Ec/I0. It is easy to implement, but has difficulty in dealing with dynamic load changes. In this algorithm, absolute handover thresholds are used. T_ADD and T_DROP are predetermined when dimensioning the network. More details about this algorithm can be found in [EIA/TIA/IS-95A]. Figure 4.33 shows the flow chat of the IS-95A soft handover algorithm.

Fig. 4.33. Flowchart of the IS-95A soft handover algorithm

Based on the IS-95A algorithm, several modified cdmaOne algorithms were proposed for IS-95B and cdma2000 systems with dynamic rather than fixed thresholds. In WCDMA, more complicated algorithm is used, as illustrated in Figure 10 [ETSI TR 125 922].


115

Fig. 4.33. WCDMA soft handover algorithm (UTRA)

The WCDMA soft handover algorithm can be described as follows: • If pilot_Ec/I0 > Best_ pilot_Ec/I0 – (AS_Th – AS_Th_Hyst) for a period of DT and the active set is not full, the cell is added to the active set. This is called Event 1A or Radio Link Addition. • If pilot_Ec/I0 < Best_ pilot_Ec/I0 – (AS_Th + AS_Th_Hyst) for a period of DT, then the cell is removed from the active set. This is called Event 1B or Radio Link Removal. • If the active set is full and Best_ candidate_pilot_Ec/I0 >Worst_Old_pilot_Ec/I0 + AS_Rep_Hyst for a period of DT, then the weakest cell in the active set is replaced by the strongest candidate cell. This is called Event 1C or Combined Radio Link Addition and Removal. The maximum size of the active set in Figure 10 is assumed to be two. Where pilot_Ec/I0 is the measured and filtered quantity of Ec/I0 of CPICH; Best_pilot_Ec/I0 is the strongest measured cell in the active set; Best_ candidate_pilot_Ec/I0 is the strongest measured cell in the monitored set; Worst_Old_pilot_Ec/I0 is the weakest measured cell in the active set. In the WCDMA algorithm (UTRA), relative thresholds rather than absolute threshold are used. Compare to IS-95A, the greatest benefit of this algorithm is its easy parameterisation with no parameter tuning being


required for high and low interference areas due to the relative thresholds. Differing from IS-95A, relative rather than absolute thresholds are used in the UTRA algorithm. Figure 4.34 shows the flow chat of the UTRA soft handover algorithm.

Fig. 4.34. Flowchart of the UTRA soft handover algorithm iii) Features of soft handover

Compared to the traditional hard handover, soft handover shows some obvious advantages, such as eliminating the “ping-pong” effect and smoothing the transmission (there is no break point in soft handover). No “ping pong” effect means lower load on the network signalling and with soft handover, there is no data loss due to the momentary transmission


117

break that happens in hard handover. Apart from handling mobility, there is another reason for implementing soft handover in CDMA; together with power control, soft handover is also used as an interference-reduction mechanism. Figure 4.35 shows two scenarios. In the top one, shown as (a), only power control is applied; in the lower one, shown as (b), power control and soft handover are both supported. Assume that the mobile is moving from BS1 towards BS2. At the current position, the pilot signal received from BS2 is already stronger than that from BS1. This means BS2 is “better” than BS1.

Fig. 4.35. Interference-reduction by SHO in UL

In (a), the power control loop increases the mobile transmit power to guarantee the QoS in the uplink when the mobile moves away from its serving BS, BS1. In (b), the mobile is in soft handover status: BS1 and BS2 both listen to the mobile simultaneously. The received signals, then, are passed forward to the RNC for combining. In the uplink direction, selection combining is used in soft handover. The stronger frame is selected and the weaker one is discarded. Because BS2 is “better” than BS1, to meet the same QoS target, the required transmit power (shown in blue) from the mobile is lower compared to the power (shown in pink) needed in scenario


(a). Therefore, the interference contributed by this mobile in the uplink is lower under soft handover because soft handover always keeps the mobile linked to the best BS. In the downlink direction, the situation is more complicated. Although the maximum ratio combining gives macrodiversity gain, extra downlink channels are needed to support soft handover. iv) Summarising the features of soft handover: A. Advantages

• Less the “ping-pong” effect, leading to reduced load on the network signalling and overhead. • Smoother transmission with no momentary stop during handover. • No hysteresis margin, leading to lower delay being equivalent to “instantaneous” macroscopic selection diversity. • Reduced overall uplink interference, leading to better communication quality for a given number of users and more users (greater capacity) for the same required QoS. • Fewer time constraints on the network. There is a longer mean queuing time to get a new channel from the target BS, this helps to reduce the blocking probability and dropping probability of calls. B. Disadvantages

• More complexity in implementation than hard handover • Additional network resources are consumed in the downlink direction (code resource and power resource) v) Simulation of a TETRA Soft-Handover Algorithm

In a CDMA based TETRA system, the mobile station is designed to support the soft-handover procedure. CDMA channel hardware and one base station to mobile telephone switching office (MTSO) trunk are occupied by the extra path in the soft-handover. We use the term pilot to refer to the pilot channel associated with the forward CDMA traffic channel. The mobile station acts as an assistant to the base station in the soft-handover process by measuring and reporting to the base station the strength of the received pilots. The procedures of the soft-handover algorithm include the maintenance of the active set based on the measured pilot strength. With a reference pilot identified in the active set, handover decision is made by comparing the strength of the pilot examined to that of the reference. Three parameters must be specified in the algorithm: the add threshold, and drop timer. If the level of any pilot that is not a member of


119

the active set exceeds the add threshold, Tadd , the pilot will be added to the set. If the pilot inside the active set drops below the drop threshold, Tdrop , the mobile station shall start a drop timer. The mobile station shall reset and disable the timer if the level of pilot goes above the drop threshold before the timer expires. The pilot is removed from the active set if the timer expires. The level of the pilot is the relative level to the reference pilot. Both Tadd and Tdrop are negative values. High thresholds and long drop timers settings tend to keep more base stations in the active set. The number of set updates will then be smaller, and the voice quality will be better because of the higher diversity gain available. On the other hand, low thresholds and short drop timer’s settings would increase the rate at which the active set updates, but save the network resources by having smaller number of pilots in the set. However, the active set update requires signaling between the mobile and base station, which can be counted as another kind of system load. Although our scope does not include a quantification of this extra signaling load, the methodology provides input to such a study. We concentrate on studying the performance of the soft-handover algorithm under the selection of different sets of parameters. The system overhead is evaluated by the base station utilization factor and the number of active set updates. The voice quality is represented by the outage probability. The definition for these can be found in the analysis of the following section. The objective of our simulation is not to model every detail in the IS-95 standard, but rather develop applicable analysis methodology which can be extended to different algorithms. Some differences between this model and IS-95 are as follows. 1. We use pilot strength while IS-95 uses Ec / I o -the ratio of pilot chip energy to total interference spectral density (actually over tracked multipaths ). 2. We only consider two base stations and an active set. IS-95 considers more than two base stations and has active, candidate, and neighbour sets. Consequently, our definitions of Tadd and Tdrop are not synonymous with those IS-95. 3. We use relative thresholds while IS-95 also uses absolute signals in addition to relative.

120 4 Channel Assignment and Multiple Access in Trunking Radio Systems [1] Model Analysis

The basic system consists of two TETRA base stations A and B separated by a distance of D. The propagation environment is characterized as log linear path loss plus shadow fading. The mobile station moves from TETRA base station A to B along the straight line connecting them. We assume there is no correlation between the received pilot strength from A and that from B. The received signal at the mobile unit from BS to A at distance d is modeled as

sa (d ) = K1 − K 2 log d + u (d )(dB)

(4.43)

where u (d ) is zero-mean Gaussian process with exponential correlation function as follows :

E{u (d )u (d + ∆d )} = σ 2 exp(∆d / d 0 )

(4.44)

d 0 is a constant to represent how fast the correlation decays with the distance increasing. The pilot strength received from base station B, sb (d ) can be expressed by a similar equation:

sb (d ) = K1 − K 2 log d + υ (d )(dB) , where υ (d ) is a zero- mean Gaussian process having the same correlation function as that of u (d ) . u (d ) and υ (d ) are independent processes. We assume the pilot signal strengths are measured by the mobile station at the end of every d s distance interval. The sampled versions of the pilot strength processes are defined as sa (n) and sb (n) , where n = 1, 2,...,[ D / d s ] . The difference between these two processes is defined as x(n) = sa (n) − sb (n). (4.45) The analysis of the handover algorithm can be solved using recursion. First, we compute the probabilities that a base station is added or dropped from the active set at the end of the d s interval assuming the mobile moving at a constant speed υ . Let us define PA→ AB (n) , PB → AB (n) , PAB → A (n) and PAB → B (n) as the probabilities that base station B is added, B is dropped and A is dropped. These probabilities can be computed using the following:

PA→ AB (n) = Pr{− x(n) > Tadd | A, − x(n − 1) < Tadd }

(4.46)


121

≈ Pr{− x(n) > Tadd | − x(n − 1) < Tadd }

PB → AB (n) ≈ Pr{x(n) > Tadd | x(n − 1) < Tadd }

(4.47)

PAB → A (n) ≈ Pr{− x(n − M ) < Tdrop | − x(n − M − 1) > Tdrop } •

n

∏

k = n − M +1

Pr{− x(k ) < Tdrop | − x(k − 1) < Tdrop } (1)

(4.48)

Where M = [υ t / d s ] and t is the drop timer expiration setting. In the above, certain conditions events, considered subsidiary, are dropped. The approximation will be compared with simulations. The last two equations use a chain-product approximation for a joint probability. Since reasonable accuracy was obtained with this simplification, it was retained. For more on calculating durations of excursions below levels, see [75]. Once these transition probabilities are found, the assignment probabilities can be calculated straightforwardly

PA (n) = PAB (n − 1) PAB → A (n) + PA (n − 1)[1 − PA→ AB (n)] PB (n) = PAB (n − 1) PAB → B (n) + PB (n − 1)[1 − PB → AB (n)]

(4.49) (4.50)

PAB (n) = PA (n − 1) PA→ AB (n) + PB (n − 1) PB → AB (n) (4.51)

+ PAB (n − 1)[1 − PAB → A (n) − PAB → B (n)] where PA (n) , PB (n) and PAB (n) are the probabilities that active set contains A only, B only , or both A and B. The initial condition is PA (n) = 1 , PB (0) = 0 , PAB (0) = 0 , Given the condition that the mobile starts moving from A toward B. Now it is possible for us to evaluate the soft-handover performance. The following parameters will be used as the performance indicators. 1. NOBS is defined as the expected number of base stations in the active set along the entire path, namely, starting with A and ending at B. 2. NOupdate is defined as the expected number of active updates for this path.


3. Poutage(n) is the outage probability

that the received signal

strength at the mobile is below the minimum quality threshold Tq . In [67], there was a transparent two-dimensional (2-D) tradeoff between the number of handovers and handover delay (or probability of outage). Here, there are three performance measures, and an optimization must be based on the costs associated with all three. These performance parameters are given by

NOBS =

1 N

N

∑ [ P (k ) + P ( k ) + 2 P k =1

A

B

AB

(k )]

(4.52)

Where N is the total number of sampling points for the path N

NOupdate = ∑ PA (k − 1) PA→ AB (k ) + PB (k − 1) PB → AB (k ) k =1

(4.53)

+ PAB (k − 1)[ PAB → A (k ) + PAB → B (k )] Poutage(n) ≈ PA (n) Pr{sa (n) < Tq | − x(n) < Tadd } + PB (n) Pr{sb (n) < Tq | x(n) < Tadd }

(4.54)

+ PAB (n) Pr{sa (n) < Tq }Pr{sb (n) < Tq } Pr{sa (n) < Tq | − x(n) < Tadd } ∞

=

⎡ y − Tadd − µ a ( d n ) ⎤ ⎡ T − µα ( d n ) ⎤ −Q⎢ q ⎥ ⎥} pb ( d n )( y ) dy σa σa ⎦ ⎣ ⎦ ⎡ −T − µ x ( d x ) ⎤ Q ⎢ add ⎥ σx ⎣ ⎦

∫ {Q ⎢⎣

−∞

(4.55)

where it is shown in (4.55): with Q( ) the standard Q function: µ a and µb are means and pb is the pdf , of the signals. In order to test the performance of our model analysis, we simulated the algorithm based on the equations above. The parameters we chose are as follows: 1. Propagation constants K1 = 0 , K 2 = 30 and the exponential constant of log linear path loss equal to three;


123

2. Standard deviation of lognormal fading σ = 8 dB; 3. Correlation constant d 0 = 20 m;

Probability

4. Sampling distance d s = 1 m. The results obtained from analysis are shown in Figure 4.36. To calculate the base station assignment probabilities, the add threshold is chosen as -1 dB, the drop threshold is -2 dB, and the timer we used is five.

1,0

0,8

PATadd1 PBTadd1 PABTadd1

0,6

0,4

0,2

0,0 0

500

1000

1500

Inter Base Station distance

Fig. 4.36. Assignment Probabilities vs InterBase Station distance

2000


Probability

0,25

PoutTadd1 PoutTadd5 PoutTadd10

0,20

0,15

0,10

0,05

0,00 0

500

1000

1500

2000

InterBase Station Distance Fig. 4.37. Outage Probabilities vs. InterBase Station Distance

Figure 4.37 shows the outage probabilities of our simulation analysis for various Tadd thresholds, i.e. for Tadd = -1 dB, -5 dB and -10 dB, respectively. The signal strength threshold in computing outage probabilities is chosen as -91 dB, under which the signal is assumed as unacceptable. Figure 4.38 illustrates the effect of the base station Tadd thresholds in the number of the BS that belong to the active set. According to this figure, a very low threshold increases the number of BSs that dedicate a channel to a TETRA terminal, thus, the available resources become even less. Finally, in figure 16 the number of the active set updates for various Tadd thresholds is presented. One can easily obtain that there is a trade-off between figures 4.39 and 4.40 as a higher threshold produces more updates to the active set of BSs, which leads to increased computational time and power in the system MSC. Thus, a “balanced” value for the Tadd thresholds should be chosen in order to satisfy both demands. A dynamic function for the suitable determination of these thresholds is the scope of our future work.


NOBS's

1,60 1,55 1,50 1,45 1,40 1,35 1,30 1,25 -10

-8

-6

-4

-2

0

Tadd thresholds Fig. 4.38. Number of BS’s vs Tadd Thresholds

NO upadates

20 18 16 14 12 10 8 -10

-8

-6

-4

-2

Tadd Thresholds Fig. 4.39. Number of Active Set updates vs Tadd Thresholds

0

125


References 1. I.Katzela and M.Naghshineh, “Channel Assignment Schemes for Cellular Mobile Telecommunications: {A} Comprehensive Survey”,IEEE Personal Communications, pp10-31,1996 2. William C. Jakes. Microwave Mobile Communications. IEEE Press. 3. W.C.Y.Lee. New Cellular Schemes for Spectral Efficiency . IEEE Transactions on Vehicular Technology, VT-36:188-192, 1987. 4. Jens Zander. Asymptotic Bounds on the Performance of a Class of Dynamic Channel Assignment Algorithms. IEEE Journal on Selected Areas in Communications, 11:926-933, 1993. 5. J.C-I. Chuang. Performance Issues and Algorithms for Dynamic Channel Assignment. IEEE JSAC, 11, 6:-, 1993. 6. Yoshihiko Akaiwa and Hidehiro Andoh. Channel Segregation-A self Organized Dynamic Allocation Method: Application to TDMA/FDMA Microcellular System. Journal on Selected Areas in Communication, 11:949-954, 1993. 7. Lewis Anderson. A Simulation Study of Some Dynamic Channel Assignment Algorithms in High Capacity Mobile Telecommunications System. IEEE Transactions on Vehicular Technology., VT-22:210-217, 1973 8. N. Bambos, S. Chen, and G. Pottie. Radio Link Admission Algorithms for Wireless Networks with Power Control and Active Link Quality Protection. Proceedings of Infocom’95, Boston, MA, April 1995. 9. R. Beck and H. Panzer. Strategies for Handover and Dynamic Channel Allocation in Micro-Cellular Mobile Radio Telephone Systems. IEEE VTC, 1:178-185, 1989. 10. B.Eklundh. Channel Utilization and Blocking Probability in a cellular mobile system with direct reentry. IEEE Transactions on Communications, 34:329-337, 1986. 11. D. C. Cox and D. O. Reudink. Dynamic Channel Assignment in High-Capacity Mobile Communications Systems. The Bell System Technical Journal, 50: 1833-1857, 1971. 12. D. C. Cox and D. O. Reudink. Dynamic Channel Assignment in Two Dimension Large-Scale Mobile Radio Systems. The Bell System Technical Journal, 51: 1611-1628, 1972. 13. Donald Cox and Douglas O. Reudink. Increasing Channel Occupancy in Large Scale Mobile Radio Systems: Dynamic Channel Reassignment. IEEE Transactions on Vehicular Technology, VT-22:218-222, 1973. 14. Joel S. Engel and Martin Peritsky. Statistically-Optimum Dynamic Server Assignment in Systems with Interfering Servers. IEEE Transactions on Vehicular Technology, VT-22:203-209, 1973.

References

127

15. A. S. Acampora and M. Naghshineh. Control and Quality of Service Provisioning in High-Speed Micro-cellular Networks. IEEE Personal Communications Magazine, 1, 1994. 16. Kazunori Okada and Fumito Kubota. On Dynamic Channel Assignment in Cellular Mobile Radio Systems. IEEE International Symposium on Circuits and Systems, 2:938-941, 1991. 17. Kazunori Okada and Fumito Kubota. On Dynamic Channel Assignment in Cellular Mobile Radio Systems. IEEE International Symposium on Circuits and Systems, 2:938-941, 1991. 18. Scott Jordan and Asad Khan. Optimal Dynamic Allocation in Cellular Systems. Submitted for publication, 1993. 19. Ming Zhang. Comparisons of Channel Assignment Strategies in Cellular Mobile Telephone Systems. IEEE Transactions on Vehicular Technology, VT-38:211-215, 1989. 20. Ming Zhang and Tak-Shing Yum. The Non-Uniform Compact Pattern Allocation Algorithm for Cellular Mobile Systems. IEEE Transactions on Vehicular Technology, VT-40:387-391, 1991. 21. M.Sengoku, M.Kurata, and Y.Kajitani. Applications of Re-arrangement to a Mobile Radio Communication System. 22. David Everitt and N.W.MacFadyen. Analysis of Mul-ticellular Mobile Radio Telephone Systems with Loss. BT Technical Journal, 2:37-45, 1983. 23. A.Gamst. Some Lower Bounds for a Class of Frequency Assignment Problems. IEEE Transactions on Vehicular Technology, 35, 1986. 24. K.Sallberg et all. Hybrid Channel Assignment and Reuse Partitioning in a Cellular Mobile Telephone System. 37th IEEE VTC, pages 405-411, 1987. 25. R.W.Nettleton. A High Capacity Assignment Method for Cellular Mobile Telephone Systems. 39th IEEE VTC, pages 359-367, 1989. 26. Mutsumu Serizawa and David Goodman. Instability and Deadlock of Distributed Dynamic Channel Allocation. Proceedings of the 43rd IEEE Vehicular Technology Conference, pages 528-531, 1993. 27. Y. Furuya and Yoshihiko Akaiwa. Channel Segregation, a Distributed Channel Allocation Scheme for Mobile Communication Systems. IEICE Transactions, 74:1531-1537, 1991. 28. V.H.MacDonald. AMPS: The Cellular Concept. Bell Sys. Tech. Journal, 58,1, 1979. 29. V.Prabhu and S.S. Rappaport. Approximate Analysis of Dynamic Channel Assignment in Large Systems with Cellular Structure. IEEE Transactions on Communications,22:1715-1720, 1974. 30. W.C.Jakes Jr. Microwave Mobile Communications, New York, Wiley, 1974. 31. R. Guerin. Queueing Blocking System with Two Arrival Streams and Guard Channels. IEEE Transactions on Communications, 36:153-163, 1988.

128 4 Channel Assignment and Multiple Access in Trunking Radio Systems [1] 32. Tomson Joe Kahwa and Nicolaos Georganas. A Hybrid Channel Assignment Scheme in Large-Scale Cellular-Structured Mobile Communication Systems. IEEE Transactions on Communications, COM-26:432-438, 1978. 33. Donald Cox and Douglas Reudink. Increasing Channel Occupancy in Large Scale Mobile Radio Systems: Dynamic Channel Reassignment. IEEE Transactions on Communications, 21:1302-1306, 1973. 34. Jun Tajima and K. Imamura. A Strategy for Flexible Channel Assignment in Mobile Communication Systems. IEEE Transactions on Vehicular Technology, VT-37:92-103, 1988. 35. Privin Johri. An Insight into Dynamic Channel Assignment in Cellular Mobile Communication Systems. European Journal of Operational Research, 74:7077, 1994. 36. I.Panoutsopoulos, S.Kotsopoulos, C.Ioannou and S.Louvros, “Priority technique for optimising handover procedure in personal communication systems”, Electronics Letters, 30th March 2000, Vol.36, No 7. 37. K.G.Ioannou, S.A.Louvros, I.C.Panoutsopoulos, S.A.Kotsopoulos, and G.K.Karagiannidis, Μember, IEEE, “Optimizing the handover call blocking probability in cellular networks with high speed moving terminals”, IEEE Communications Letters, Vol 6, No. 10, October 2002. 38. K.Ioannou, S.Kotsopoulos, P.Stavroulakis, “Optimizing the QoS of High Speed Moving Terminals in cellular networks” International Journal of Communications Systems 2003; 16:851-863 (DOI: 10.1002/dac621). 39. K.Ioannou, I.Panoutsopoulos, S.Koubias and S. Kotsopoulos, “A new Dynamic Channel Management Scheme to increase the performance index of Cellular Networks”, Electronics Letters, 10th June 2004, Vol.40, No 12, pp 744-746. 40. I.Panoutsopoulos, S. Kotsopoulos, K.Ioannou, S. Louvros, “Αn Alternative Queuing Technique to Optimize the Performance of TDMA Based Mobile Communication Systems”, Wireless Personal Communications, Volume 31, Issue 3 - 4, Dec 2004, Pages 149 – 159. 41. E.Dimitriadou, K.Ioannou, I.Panoutsopoulos, S.Mougiakakou, P.Stavroulakis and S.Kotsopoulos, “An Alternative Dynamic Channel Assignment Technique for use in Satellite-Aided Cellular Systems”, WSEAS Transactions on Communications, Issue 7, Volume 4, July 2005, p 345-354. 42. E.Dimitriadou, K.Ioannou, I.Panoutsopoulos, A.Garmpis, S. Kotsopoulos, “Priority to Low Moving Terminals in TETRA Networks”, WSEAS Transactions on Communications, Issue 11, Volume 4, November 2005, pp 1228-1236. 43. D.Karaboulas, S.Louvros, K.Ioannou, I.Panoutsopoulos and S.Kotsopoulos, “A two layer HASP Architecture solution for optimizing the performance subscribers in microcellular networks”, WSEAS Transactions on Systems, Issue 11, Volume 4, November 2005, pp 1821-1828. 44. K.Ioannou, E.Dimitriadou, A.Ioannou, I.Panoutsopoulos, A.Garmpis and S. Kotsopoulos, “Efficient Channel Assignment Technique providing Priority to Emergency Calls in Wireless Cellular Security Networks, WSEAS Transactions on Communications, Issue 1, Volume 5, January 2006, pp 84-91.

References

129

45. K.Ioannou, E.Dimitriadou, A.Ioannou, I.Panoutsopoulos and S.Kotsopoulos, “Hierarchical Dynamic Channel Assignment Technique”, WSEAS Transactions on Communications, Issue 1, Volume 5, January 2006, pp 115-122. 46. B.Vassiliadis, J.Tsaknakis, L.Drossos, S.Sirmakessis, K.Ioannou, “Providing Service Standardisation for Virtual Enterprises” WSEAS Transactions on Circuits and Systems, Issue 11, Volume 4, November 2005, pp 1545-1552. 47. K.Ioannou, J.Pylarinos, S.Louvros, I.Panoutsopoulos, I.Kougias, A.Garmpis and S.Kotsopoulos, “A New Tree Based Call Admission Control for Wireless Networks”, WSEAS Transactions on Systems, Issue 3, Volume 5, March 2006, pp 352-357. 48. J.Pylarinos, S.Louvros, K.Ioannou, A.Garmpis and S.Kotsopoulos, “Finite Storage Queues for Wireless Cellular Networks”, WSEAS Transactions on Communications, Issue 2, Volume 5, February 2006, pp 218-224. 49. L.Raptis, D.Kagklis, Y.Patikis, G.Hatzilias, N.Zervos, T.Zahariadis, D.Crhonis, K.Ioannou, A.Garmpis, “Building a Multimedia Next Generation Network”, WSEAS Transactions on Communications, Issue 2, Volume 5, February 2006, pp 225-230. 50. G. Agapiou, D.Vali, D.Charalambakis, P.Georgiadis, E.Plakidis, P.Rorris, K.Ioannou, A.Garmpis, “Experimental Performance Evaluation of the emerging WiMAX technology”, WSEAS Transactions on Information Science and Applications, Issue 2, Volume 3, February 2006, pp 322-327. 51. G.Agapiou, T.Doukoglou, P.Georgiadis, G.Georgakopoulos, E.Plakidis, P.Rorris, K.Ioannou, A.Ioannou, A.Garmpis, “Throughput and interference measurements in copper wires of xDSL type modems”, WSEAS Transactions on Information Science and Applications, Issue 2, Volume 3, February 2006, pp 352-357. 52. N. Deligiannis, S. Louvros, K.Ioannou, S. Kotsopoulos, « Location Positioning Techniques Investigation in GSM Networks », WSEAS TRANSACTIONS on COMMUNICATIONS, Issue 6, Volume 5, June 2006, ISSN 1109-2742, pp 1159-1168. 53. K.Ioannou, I.Panoutsopoulos and S.Kotsopoulos, “Optimizing the QoS of Ultra High and High Speed Moving Terminals in Satellite-aided cellular networks”, Wireless Personal Communications, Volume 39, Number 3, November 2006, pp307-319”. 54. Ι. Panoutsopoulos, S. Kotsopoulos, C. Ioannou and S. Louvros, “Priority technique for optimizing handover procedure in personal communication systems”, Electronics Letters, 30th March 2000, Vol.36, No 7. 55. T.S. Rappaport, Wireless Communications (Prentice-Hall, Englewood Cliffs, NJ, 1996). 56. M.F. Catedra and J.P.-Arriaga, Cell Planning for Wireless Communications (Artech House, 1999). 57. N. Passas, S. Paskalis, D. Vali and L. Merakos, Quality-of-Service oriented medium access control for wireless ATM networks IEEE Communications Magazine (November 1997) 42–50.

130 4 Channel Assignment and Multiple Access in Trunking Radio Systems [1] 58. M. Naghshineh and M. Willebeck-LeMair, End-to-end QoS provisioning in multimedia wireless/mobile networks using an adaptive framework, IEEE Communications Magazine (November 1997) 72–81. 59. L. Hu and S.S. Rappaport, Personal communication systems using multiple hierarchical cellular overlays, IEEE Journal on Selected Areas in Communications 13(2) (February 1995) 406–415. 60. R. Coombs and R. Steele, Introducing microcells into macrocellular networks: A case study, IEEE Transactions on Communications 47(4) (April 1999) 568–576. 61. T. S. Rappaport, “Wireless Communications”, July 1999, ISBN 0-13-375536-3. 62. D.Hong and S.S. Rappaport: “Traffic model and performance analysis for cellular mobile radio telephone systems with prioritized and non prioritized handoff procedures”, IEEE Trans., 1986, VT-35, pp 77-92. 63. Kawabata, K., T. Nakamura, and E.Fukuda, “Estimating velocity using diversity reception,” Proc. IEEE VTC 1994, Stockholm, Sweden, 1994, pp. 371-374. 64. J. Zander, S-L Kim, “Radio Resource Management for Wireless Networks”, 2001, ISBN1-58053-146-6. 65. B.L.Lim and Lawrence W.C. Wong, “Hierarchical Optimization of Microcellular Call handoffs”, IEEE Trans. Vech. Vol. 48, no.2, pp.459-466, March 1999. 66. Katzela, I. Naghshineh, M., “Channel assignment schemes for cellular mobile Telecommunication systems”, IEEE Personal Communications, June 1996 Volume: 3, Issue: 3, p 10-31,ISSN: 1070-9916 67. R. Vijayan and J. M. Holtzman, “A model for analyzing handoff algorithms,” IEEE Trans. Veh. Technol., vol. 42, pp. 351–356, Aug. 1993. 68. N. Zhang and J. M. Holtzman, “Analysis of handoff algorithms using both absolute and relative measurements,” in 44th IEEE Vehicular Technology Conf., Stockholm, Sweden, June 1994, pp. 82–86. 69. S. Wang and I. Wang, “Effects of soft handoff, frequency reuse and nonideal antenna sectorization on CDMA system capacity,” in 43rd IEEE Vehicular Technology Conf., Secaucus, NJ, May 1993, pp. 850–854. 70. M. Wallace and R. Walton, “CDMA radio network planning,” in 3rd Int.Conf. Universal Personal Communications, San Diego, CA, Oct. 1994, pp. 62–67. 71. A. J. Viterbi et al., “Soft handoff extends CDMA cell coverage and increase reverse link capacity,” IEEE J. Select. Areas Commun., vol. 12, no. 8, 1994, pp. 1281–1288. 72. M. D. Austin and G. L. Stuber, “Cochannel interference modeling for signal strength based handoff analysis,” Electron. Lett., vol. 30, no. 23, pp. 1914, 1994. 73. K. M. Rege, S. Nanda, C. F. Weaver, and W.-C. Peng, “Analysis of fade margins for soft and hard handoffs,” in PIMRC’95, Sept. 1995, pp. 829–835. 74. K.Ioannou, E.Dimitriadou, A.Ioannou, I.Panoutsopoulos and S.Kotsopoulos, “Hierarchical Dynamic Channel Assignment Technique”, WSEAS Transactions on Communications, Issue 1, Volume 5, January 2006, pp 115-122.

References

131

75. P. C. Chen, N. Mandayan, and J. M. Holtzman, “Minimum duration outage for cellular systems with reception diversity,” in ICUPC’96, Sept. 1996, pp. 245–249. 76. S. Agarwal, J. M. Holtzman, “Modelling and analysis of handoff algorithms in multi-cellular systems”, Vehicular Technology Conference, 1997 IEEE 47th, Vol. 1, pp. 300-304, May 1997. 77. A. S. Anpalagan and I. Katzela, “Overlaid cellular system design with cell selection criteria for mobile wireless users”, Electrical and Computer Engineering, 1999 IEEE Canadian Conference on , Vol. 1 , pp: 24 -28, May 1999. 78. F. Ashtiani, J. A. Salehi, M. R. Aref, M. Nasiri-Kenari, “A new soft-handoff management algorithm with two decision boundaries,” Personal, Indoor and Mobile Radio Communications, 2001 12th IEEE International Symposium on, Vol. 1, pp. 54-58, Sept. 2001. 79. X. H. Chen, “Adaptive traffic-load shedding and its capacity gain in CDMA cellular systems,” Communications, IEE Proceedings-, Vol. 142, Issue. 3, pp. 186-192, June 1995. 80. A. Chheda, “A performance comparison of the CDMA IS-95B and IS-95A soft handoff algorithms,” Vehicular Technology Conference, 1999 IEEE 49th, Vol. 2, pp. 1407-1412, 1999. 81. J. W. Chang and D. K. Sung, “Adaptive channel reservation scheme for soft handoff in DS-CDMA cellular systems,” Vehicular Technology, IEEE Transactions on, Vol. 50, Issue. 2, pp. 341-353, March 2001. 82. W-Y. Choi, K. Yeo and C-H Jun, “Soft handoff algorithm with pilot signal thresholds adjustment in CDMA cellular systems,” TENCON 99. Proceedings of the IEEE Region 10 Conference , Vol.1, pp. 498-501, 1999. 83. B. Homnan, V. Kunsriruksakul, W. Benjapolakul, “A comparative performance evaluation of soft handoff between IS-95A and IS-95B/cdma2000,” IEEE APCCAS’2000, pp. 34 –37, 2000. 84. S-H. Hong, S-L. Kim, H-S. Oh, C-E. Kang and J-Y Son, “Soft handoff algorithm with variable thresholds in CDMA cellular systems”, Electronics Letters, Vol. 33, Issue. 19, pp. 1602-1603, 1997. 85. S. J. Hong and I. T. Lu, “Soft Handoff Algorithm Parameter Optimisation in Various Propagation Environments”, IEEE VTC’01, pp. 2549-2553, 2001

5 Video Transmission over TETRA

Ilias Politis, Michail Tsagkaropoulos, Stavros Kotsopoulos

5.1 Introduction TETRA provides efficient and secure group voice and data communication. In addition to transactional data solutions for public safety, it also supports wireless browsing, e-mail, vehicle location and fingerprint resolution. Although, visual information such as maps, digital pictures and slow motion video can be transmitted over 2G/3G cellular digital radio networks, but these are too slow when transmitted over current TETRA data. If speeds were increased, TETRA could be used to offer completely new data applications for public safety, for example, remote surveillance and recognition in real time, fast checking of identities, monitoring cargos and containers and sending digital maps of buildings. The most promising way of increasing the speed of TETRA communications is through the use of TETRA Enhanced Data Service (TEDS) [5]. This is an enhanced TETRA air interface standard that increases data speeds over TETRA up to 50-300kb/s, over 10 times faster than Multi-slot Packet Data over TETRA. TEDS clearly fulfils all the mandatory public safety high-speed data requirements, whereas public cellular technologies, such as GSM or CDMA data services, fall short of meeting the reliability requirements of Public Safety in emergency situations. In the last years, there is an increasing attention among the researchers towards the transmission of multimedia applications and services over heterogeneous wireless networking technologies. Additionally, the research effort for the and 3G/4G vision of interworking among heterogeneous technologies to achieve video session continuity, retain video QoS characteristics etc., amplifies the need to evaluate the conditions and restrictions under which seamless video session continuity can be achieved [13]. Interworking between WLAN and TETRA networks is currently under research by the international scientific community. Such interworking may

134


necessitate roaming partnership, billing and accounting, retaining QoS, etc. The issue of video transmission over TETRA and TEDS networks has been addressed by a large number of researchers. It is apparent that video transmission within the TEDS network may be subject to QoS degradation due to bandwidth limitation when supporting large number of users [14]. Hybrid networks can support large number of users and in the same time guarantee QoS due to the ability to handover traffic towards a network with higher available capacity. Within the context of this article, we monitor the QoS regarding MPEG-4 video streaming traffic delivery, in terms of both packet loss and perceived image quality, over TEDS networks. The rest of the article is organized as follows. In Section 5.2, we attempt to follow the evolution in public safety mobile networks and the role of TETRA networks. The need for more complex context in the information exchange, guaranteed quality of service and secure, flexible and scalable infrastructures led to the standardization of TETRA and TEDS systems. Section 5.3, includes an extensive overview of data transmission over TETRA, followed by an insight on the evolutions incorporated in TEDS standard that allowed video and high data rates support. A detailed analysis of the current video encoding techniques and error concealment methods are contained in Section 5.4. Particular interest is given to MPEG-4 encoding standard as it enables higher video compression rates, thus making it an ideal solution for video traffic over TETRA and TEDS networks. The main two techniques of video encryption are described in the section as well. In the following Section 5.5, we provide a performance analysis of video transmission over TEDS network. The perceived QoS assessment based on specific network condition and traffic loads provides a performance indication of video traffic over TETRA. The current and future scenarios and trends for TETRA integration with other Cellular and IP-based networks are indicated in Section 5.6. In this section a very interesting scenario of TETRA-WLAN integration is also studied. Finally, the article is concluded in the last Section 5.7.

5.2 Evolution of Public Safety Mobile Networks The mobile communications requirements of the public safety and disaster recovery (PSDR) organizations are continuously evolving. Traditional voice and data services support is no more sufficient for the particularly demanding functional and performance requirements of modern PSDR communication systems. Currently, the information exchanges between the control centre and the dispatched on-site units and directly

5.2 Evolution of Public Safety Mobile Networks

135

between on-site units are characterized by complex context (voice, data and real time video services, highly mobile end terminals, etc.), very high level quality of service (QoS) and end-to-end security [10]. The Professional Mobile Radio (PMR) communication systems and especially the Terrestrial Trunked Radio (TETRA) network are extensively utilized by law enforcement, fire, ambulance, search and rescue agents, government administration and so forth. In the current section we investigate how the ever evolving and multiplying application and service demands of public safety mobile communication networks fuels the transition from the traditional PMR systems to TETRA and TETRA Release 2 [4] (Enhanced Data Services – TEDS) solutions. Professional or Private Mobile Radio occupies a small fraction of the mobile communications market, mainly because its aim was and is to function as a professional tool for a wide range of corporate mobile business communications people and ‘group’-oriented users that include PSDR organizations and other commercial users such as transport and taxi companies [23]. They use PMR intensively because both its unique functionality and its performance match their requirements better than other mobile technologies. The key functionalities of PMR are: • push-to-talk, release-to-listen operation over a single radio channel. • point-to-multipoint communications • large coverage areas • closed user groups The PMR market has traditionally been scattered in many dimensions in terms of technologies, frequency allocation etc. The first international standardization effort resulted to the introduction of the analogue MPT 1327 trunked radio standard, that lead to a market success in most parts of the world. However analogue PMR systems could not meet the needs and performance requirements of modern public safety communication networks. Their major deficiencies include very low capacity, no available data facilities and inadequate security provisioning. 5.2.1 Evolving Data services for public safety Consumers and public safety operatives have different needs when it comes to data services. Critically, public safety personnel demand that the service is available during exceptional situations and major incidents [6]. Data services are required to work in remote locations, possibly outside the coverage of public cellular networks, such as on mountainsides, at sea, deep within forests, in the desert – wherever the incident may have taken place. Key features of the service are geographic rather than demographic

136


radio coverage, provision of a PMR network designed to guarantee GoS and to meet peak loads in times of emergency and a highly flexible design. The services must also work when public cellular networks are congested or may be out of operation due to power failures or other malfunctions. In terms of Authorization and security the unauthorized access to public safety networks must be prevented. Security levels can range from no encryption up to the strongest end-to-end encryption. Data services for public safety must be scaleable, work seamlessly across organizational and geographical boundaries enabling high levels of interoperability between high and lower speed data, as well as between networks used to access the services, hence more co-ordinated response to major incidents [22]. The data services need to support at least vehicular mobility and provide handovers between neighbouring cells and service areas without dropping connections. The access to both voice and data services, must be fast. Support for legacy data applications that use today’s lower speed TETRA data, GPRS, analogue network radio modems or specific digital data radio networks, is also a requirement. 5.2.2 The TETRA solution to PSDR communication environment Unlike traditional analog PMR systems, TETRA created by ETSI incorporates efficient data communication capabilities. The system is used throughout the world, foremost in the public safety sector and by various government institutions [16],[17]. This new standard provides higher level QoS, flexibility and improved security of communications. It is interesting to notice that the standards for TETRA have been developed at a time when the capability of GSM was being greatly enhanced through developments such as GPRS, CAMEL, EDGE, etc. However, the public cellular technologies of 2G and 2.5G were not suitable for mission-critical performance parameters, such as functionality, performance and a guaranteed high level service. The most important functionalities of current PMR TETRA include: • • • • • • •

fast call set-up (300ms-500ms) group and broadcast calls despatch operations (command and control) packet data direct mode (users can communicate with each other directly) user-selectable priority levels with pre-emption for emergency calls multiple security schemes & algorithms


137

authentication of radios and networks air interface & end-to-end encryption - disabling of stolen radios • signalling during voice call -

These key features enable TETRA to open up a new global market and to fulfill the requirements of a wider than previously group of Professional Mobile Radio users. PSDR agencies in Europe were among the first users to implement TETRA, which is also the ideal choice for commercial Public Access Mobile Radio (PAMR) networks due to superior frequency efficiency, high data transfer rate and excellent connectivity possibilities to other networks among other advanced technical characteristics [17]. In more details, the TETRA standard is based on a combination of Frequency Division Multiple Access (FDMA) and Time Division Multiple Access (TDMA) techniques. Specifically, it provides four time frames on a single frequency carrier. Each time frame or slot acts as a single communication channel, thus enabling the provision of system facilities to numerous users through highly efficient use of allocated spectrum. Moreover, TETRA supports different mechanisms that can dynamically manage communication resources in order to maintain a high level of grade-ofservice (GoS) especially in crisis situations [6]. According to TETRA specifications, several types of air interface are defined. The most commonly used interface is the voice plus data (V+D) air interface [2], which is optimised for mixed (simultaneous) transfer of sound and data. The packet data optimized (PDO) air interface that is optimised for packet data transmission via radio channels. It does not support line type services and short message services. Finally, the direct mode operation (DMO) air interface [1]. The latter forms the means of enabling a fundamental mode of ad hoc networking, in which the mobile stations communicate with each other by using radio-interface for direct communication mode (mobile-to-mobile). In this mode, intercommunication of mobile stations is independent on the network, i.e. without mediation of base stations. This is particularly important in emergency situations that arise in locations outside the V+D coverage area [7]. The V+D air interface operates in trunking mode, that is, it allocates and releases the available radio resources dynamically and on demand basis. This way, the available radio spectrum can be efficiently shared across many different groups of users, or even across many different PSDR agencies. Currently in terms of services, the TETRA supports a range of voice, data, and supplementary services, including dispatch services, group and private voice services, security services, telephone interconnect services, short data and status messaging, and packet data services [20]. All of these

138


services are characterized by mission-critical performance figures such as fast call setup (typically, between 300–500 ms), fast message transmission, priority-based call handling, advanced encryption and authentication, and so forth. Moreover, several IP-based data services are used today over TETRA technology, the most common being the Automatic Vehicular Location (AVL) and the remote database query services [45]. 5.2.3 The Market Considerations A market survey carried out in 2001 aimed at identifying what enhancements the users wanted for TETRA Release 2 [4]. Figure 1, shows the relative weighted importance of the new requirements collected from the survey.

Fig. 1. Data Interworking requirements 2001

As can be seen from the survey the most important enhancement is high-speed data. However, since 2001 significant market changes have occurred resulting in the need for interworking and roaming, and SIM enhancement among others. In Figure 2, it can be seen that there is a great deal of variation between the needs of different market sectors. Only the military has not put a high importance on high-speed data. Furthermore, it can be seen that there is a high variation of the prioritization across the market segments.


139

Fig. 2. Importance of High Speed Data per market sector (source ETSI)

The feedback from the market enforced the need for enhancing existing TETRA and TETRA Release 2 [4] in order to support higher data rates with QoS provisioning. TETRA Enhanced Data Services standard has been introduced as the answer to the market demands described above. 5.2.4 TETRA Enhanced Data Service-TEDS Provisioning of wide-area coverage for high-speed data is inevitably a significant investment; the most cost-effective alternative should be preferred. The standard is being developed by ETSI under the denomination “TETRA Release 2”. ETSI has almost completed the standardization of TETRA Enhanced Data Service (TEDS) [5]. This is an enhanced TETRA air interface standard that increases data speeds over TETRA up to 50-300kb/s,

140


over 10 times faster than Multi-slot Packet Data. TEDS coverage at its highest speed is not the same as TETRA coverage, making it available only over a limited range. A dynamic link adaptation mechanism is used to connect to a lower speed service at cell boundaries, making the service still continuously available. In shared use, each radio terminal still gets the maximum possible bit-rate when there is no congestion, depending on its radio path. Sharing the infrastructure and sites with TETRA voice and low speed data, TEDS can provide the good coverage, security, availability, scalability, interoperability and QoS that is built into the public safety TETRA voice network [14],[21]. The following are some of the most important requirements that have been addressed by TEDS or will be addressed by future Releases: • high-speed packet data in support of multimedia and other high-speed applications • additional speech codecs to allow interconnection to other 3G networks without transcoding and to provide enhanced voice quality • further enhancements to the air interface to provide increased benefits in terms of spectrum efficiency, subscriber capacity, system performance, quality of service, size and cost of terminals, etc • production and/or adoption of standards to provide improved interworking and roaming between TETRA and public mobile networks, such as GSM, GPRS and UMTS • evolution of the TETRA SIM, with the aim of convergence with the universal SIM (USIM), to meet the needs for TETRA-specific services while gaining the benefits of interworking and roaming with public mobile networks, such as GSM, GPRS and UMTS. Concluding this section it has to be mentioned that TETRA specifications are constantly being evolved by ETSI and enriched with new features and in order to fulfill all the mandatory public safety high-speed data requirements, whereas public cellular technologies, such as GSM or CDMA data services, fall short of meeting the reliability requirements of Public Safety in emergency situations [21]. However, present and future TETRA networks can be significantly enhanced and complemented by mobile broadband systems. Thus, new more advanced services envisioned in the next generation of PSDR communication systems such as remote patient monitoring, two-way real-time video, 3D positioning and GIS services, etc., will be better achieved. What follows in next section is a more technical insight on how real-time video transmission over TETRA and TEDS can be achieved.

5.3 Overview of DATA Transmission over TETRA 141

5.3 Overview of DATA Transmission over TETRA The previous section has considered the evolution of private and public safety communication networks from the analogue PMR systems to TETRA and TEDS. The new standards were developed in order to support the more demanding services and applications required for the modern private mobile domain. In this section we introduce the fundamental technical characteristics of both TETRA (Voice+Data) and TEDS and the basic services and applications that each can support. The following analysis includes an overview of the architectures, protocols and services supported by TETRA (V+D) system [2]. We will investigate the circuit and packet modes as well as the data service capabilities, short data services, teleservices and supplementary services as described in ETSI specifications. Moreover, the study contains an insight in the recently developed TEDS specifications and the enhancements made on TETRA (V+D) specifications in order to meet the public safety requirements for QoS guaranteed high-speed data communication and TETRA’s integration with GPRS/3G and WLAN technologies. 5.3.1 TETRA (V+D) Technical Characteristics The different design in radio parameters between TETRA and GSM/GPRS radio parameters is due to the fact that TETRA has been designed to allow migration from analogue PMR systems. According to ETSI specifications TETRA utilizes the existing VHF and UHF PMR frequencies and TDMA technology with four user channels interleaved into one carrier with 25 kHz carrier spacing. Since the design of TETRA was driven by the enhanced needs of PMR user groups it provides a range of features like fast call set up, higher QoS, scalability and seamlessness [11]. Network Architecture

The network architecture is defined in terms of six essential interfaces that ensure an open multi-vendor market, interoperability, interworking and network management [45], as illustrated in Figure 3. • Air Interphase – ensures interoperability of different terminal equipment • Terminal Equipment Interphase (TEI) – supports indipendent development of mobile data applications • Inter-System Interphase (ISI) – for interconnecting TETRA networks from different manufactures

142


• Direct Mode Operation (DMO) [1]- ad hoc networking, in which the mobile stations communicate with each other by using radio-interface for direct communication mode

Fig. 3. TETRA Network Architecture

Also a Line Station Interface is standardized. It should be noted, that the interfaces inside the TETRA Switching and Management Infrastructure (SwMI), which is the core component that comprises the necessary networking, switching, management, and service-provision elements of the system are not standardized. This provides the essential benefits of an open market, but leaves the manufacturers the freedom to implement the most cost-efficient network solutions. Specifically, SwMI comprises of up to six system components, the mobile station, line station, direct mode mobile station, a gateway, the network management unit and the individual TETRA network. These are interconnected by using the above specified interfaces. SwMI provides a common network domain for its included components. The functional network Configuration of TETRA [45] is shown in Figure 4.


Fig. 4. TETRA SwMI functional configuration

Circuit Mode (V+D)

TETRA has been designed as a trunked system that allows simultaneous transmission of circuit switched voice and data. This mode of operation also allows the effective and economical sharing of the usage of the network between several organizations without compromising the security and privacy. In circuit mode each source is allocated a traffic channel for the duration of the call irrespective of whether that source is active. In addition to traffic channels (TCH) the physical channels (carrier frequency, time slots) provide the mechanism for transmitting logical channels. A physical channel may be used for several logical channels on a shared basis; hence it involves the concept of multiplexing [15].

Fig. 5. Frame structure

144


TETRA has been designed to operate in the 150 MHz – 900 MHz frequency range and for each cell one or more pairs of carriers for uplink and downlink have been allocated. The separation between the two is 45 MHz in UHF and 10 MHz in VHF band. TETRA utilizes the TDMA technique to divide the carrier into four slots (four physical channels) of duration 14.167 ms. The TDMA frame structure has four time slots and a period of 56.67 ms. This is further organized as 18 TDMA frames per multiframe of 1.02 s duration. The multiframe is repeated 60 times in order to produce a hyperframe of duration 61.2 s, which is related to encryption and synchronization. In circuit mode voice and data operation traffic from an 18 frame multiframe length of time is compressed and conveyed within 17 TDMA frames, thus allowing the 18th frame to be used for control signaling without interrupting the flow of data. This 18th frame is called the control frame and is the basis for Slow Associated Control Channel (SACCH). The SACCH provides the background control channel signaling that is always present, even in minimum mode when all channels are allocated to traffic. The gross bit rate of one channel is 9 kbps, into which speech is coded with 4.8 kbps net bit rate using ACELP coding. The modulation method applied in TETRA is π/4-DQPSK - a linear modulation. Table 1 summarizes the above standardized parameters of TETRA. Table 1. Main TETRA parameters Parameters Carrier Spacing Modulation Carrier Data Rate Voice Coder Rate Access Method User Data Rate Maximum Data Rate Protected Data Rate

Values 25 kHz π/4 DQPSK 36 kbps ACELP (4.56 kbps net, 7.2 kbps gross) TDMA with 4 time slots per carrier 7.2 kbps per time slot 28.8 kbps Up to 19.2 kbps

Direct Mode

In addition to the trunk mode, TETRA can support a direct mobile-tomobile communication without the need for network infrastructure. DMO can be used when the mobile station is located outside the coverage area of the network, or as a more secure communication channel within the coverage of the network [7]. Moreover, gateway functions are defined in order to essentially extend the coverage of the network and allow for communication between DMO terminals and the trunked network. Furthermore, in cases of poor or no network coverage TETRA allows the use of repeater


stations which are independent of the trunked network. The following Figure 6 illustrates the mentioned mode of operation.

Fig. 6. Direct Mode of Operation

• • • • •

The services provided by direct mode include [15]: Individual or group calls Call set-up with and without presence check Circuit mode operation with and without encryption Pre-emptive calls User defined and pre-defined short message exchange

TETRA Connectivity

The concept of connectivity between different network types (TETRA or non-TETRA) is becoming increasingly important. Particularly, TETRA networks are expected to satisfy specific organizational needs such as, serving an airport where the network will provide coverage to just a small area, or to facilitate regional emergency services, where the network coverage will be extended to a wide area. Therefore, TETRA network installation is primarily based on the specific site needs and not on providing a contiguous and complete coverage as public networks do. As a result TETRA networks facilitate a wide range of connections to external networks that can be accessed from the mobile terminal. These networks can be public or private telephone networks, different types of data networks and command or control systems. Additionally, these networks may be owned by single or different organizations. An interconnection like this allows network resource sharing at various levels and provides an integrated networking environment.

146


Fig. 7. TETRA interconnection configuration (source [15])

The interconnection between different networks is described by TETRA standards that specify the requirements for interworking over the intersystem interface (ISI). The connectivity to different networks combined with bandwidth-on-demand makes TETRA a superior platform for data application development. TETRA Network Security

TETRA networks support several mechanisms for secure communications. The technology inherently includes protection mechanisms for voice, data, signaling and user identities. In order to implement secure communication at various levels TETRA standard defines the following features [8]: • Authentication mechanisms - for the users with an identity stored in SIM - for mobile stations with a unique equipment number - for the network and the network managment system • identity confidentiality of individual or group users • data integrity and authentication of their origin for signalling • confidentiality of signalling information • secure functions for air interface key managment The protection mechanisms defined in TETRA standard are listed below: • multi-level authentication – access permissions to users • air interface encryptions – encrypts the radio path between terminal and base station


• user anonymity – with alias addressing • terminal enable/disable – by the infrastructure over the air • end-to-end encryption – for the most critical applications where encryption is required from end-to-end • frequency hopping – the TDMA slot structure permits slow frequency hopping to avoid jamming 5.3.2 TETRA Network Services The basic services that are available with TETRA are bearer services and teleservices. In more details the former characterizes services which provide information transfer between user network interfaces, excluding the functions of the terminal. TETRA bearer services are defined for data transfer. Additionally, teleservices provide complete capability for communication between end users. In TETRA standards teleservices cover voice communication services. There are also supplementary to bearer services and teleservices for very flexible system applications. A typical example of PMR type supplementary service is allocation of priority call, while call forwarding is a typical telephone type supplementary service. The following Table 2 summarizes the most common of the above services [8]. Table 2. Most common TETRA, teleservices, bearer and supplementary services TETRA Teleservices Individual Call Group Call Acknowledged Group Call Broadcast Call TETRA Bearer Services Circuit mode data Circuit mode protected data Circuit mode heavily protected data Connection oriented packet data Connectionless packet data Essential Supplementary Services Call Authorised by Dispatcher Area Selection Access Priority Priority Call Late Entry Pre-emptive Priority Call

148

5 Video Transmission over TETRA Discrete Listening Ambience Listening Dynamic Group Number Assignment

As it has been mention in the previous section there are three types of TETRA systems: V+D, PDO and DMO. These are identified in terms of services they support as follows: • V+D system - Circuit mode speech and data - Packet mode data - Short data services (SDS) over four time slots • PDO system - Packet mode data over the equivelant of four time slots in V+D system • DMO system - Circuit mode speech and data - Short data services over one time slot Analytically, circuit mode data services that are supported by TETRA V+D [2] and TETRA DMO [1] depend on the level of data error protection that is required according to the nature of the transported data and the number of channels aggregated under multi-slot operation. Applications that tolerate degradation to the QoS levels such as speech and image can be transmitted with the minimum required protection. On the other hand, real time video transmission or signaling messages would be provided with the highest level of protection. In addition to protection levels data transmission requires adequate channel capacity to be available. Capacity depends primarily on the number of channels aggregated and the protection level, further it is determined by the volume of data to be transmitted, application’s tolerance to delays and the cost of the service. Moreover, packet mode data services, which are supported by V+D and PDO systems, are characterized by their bursty nature. Thus PDO systems employ a statistical multiplexing scheme at the gross rate of 36 kbps over the whole channel provided by the carrier. In the contrary V+D system uses a TDMA scheme that is also used for digitized speech transmission. As shown in Table 2 there are two types of packet mode data services, connection and connectionless oriented. The former establishes virtual connections between the source and the destination in order to transfer the packets. The latter, on the other hand, is similar to the IP protocol and transfers a single data packet without establishing a virtual connection. Finally, SDS is datagram based service and it uses the spare capacity of the signaling channel in order to transmit messages from point-to-point or


point-to-multipoint. Therefore, no channel establishment is required and the exchange of messages occurs in parallel to ongoing connections. In order to ensure that it will meet the needs of the demanding users, the development of TETRA standard was greatly influenced by the user comments and suggestions, especially of those involved in emergency services. Hence, TETRA standard contains high functionality for PSDR services and is very well suited for commercial trunked radio users. 5.3.3 High Speed Data service provisioning TETRA Enhanced Data Service has been developed to provide a High Speed Data service in response to user needs. TEDS provide a packet data solution that is integrated with existing TETRA systems, and within the spectrum constraints of some existing TETRA users. TEDS aims to enhance the services and facilities of TETRA in order to meet the emerging user requirements for high speed and multimedia services, utilize new technologies and, by maintaining the competitiveness with other wireless technologies, increase the future proofness of TETRA as the standard for PMR and PAMR world-wide. In details, TEDS provides high-speed data approximately 10 times that available in existing TETRA multi-slot packet data, in order to support multimedia and other high-speed data applications required by existing and future TETRA users. TEDS physical layer uses a range of adaptive modulations between QAM64/QAM16/QAM4 in addition to the present π/4 DQPSK and a number of different carrier sizes from 25 kHz to 150 kHz. Moreover, Turbo Coding is used to optimize data throughput. This provides the users with a flexibility of tailoring the maximum air interface speed from the present TETRA to those provided by 3G systems. The following classes of service are provided over the air interface in TEDS implementation with associated speeds, priorities and QoS attributes [14]. • Real-time class • Telemetry class • Background class Also the required spectrum allocation can be selected to suit particular requirements. Specifically, TEDS will increase the error protected TETRA data speed to 48kb/s in a 25 kHz TETRA channel, 96 kb/s, 192 kb/s, and 288 kb/s in 50 kHz, 100 kHz, and 150 kHz channels respectively. TEDS carriers are dedicated to the high-speed data service and in practice, cannot share the same capacity with TETRA voice and slow data at the same time. Therefore, it increases both the number of transceivers and carriers by 25-50% and the data capacity, compared to multi-slot packet data.

150


TEDS introduces adaptive link control to the TETRA system whereby modulation type and coding rate can be changed adaptively to improve link performance under different propagation conditions. TEDS standard supports full backward compatibility with the current TETRA standard. Hence, it can provide all the services that the current TETRA network offers, by relaying on the current TETRA control channel, signaling and roaming. In addition, the TEDS IP service maintains backward compatibility with the existing TETRA IP service. A number of new facilities have been added to ensure that transmission of several multimedia services with speeds approaching 500 kbit/s is possible. TEDS capability to support high speed IP traffic over the air interface is based on a new physical layer design and on the modified higher protocol layers of existing TETRA systems. For the ease of compatibility TEDS uses the same control channel as the existing TETRA standard. In the following Figure 8 that illustrates the newly defined protocol stack, Um is the air interface and Gi is the IP packet mode gateway to the IP application hosts.

Fig. 8. TEDS air interface protocol stack

In addition to the TETRA V+D interfaces defined by ETSI, TEDS standard [2] proposes further interfaces to enable integration with existing cellular public, IP and TETRA networks. Figure 9 shows how such integration can be achieved by utilizing the Air Interface (Um) and the additional Packet Data Network (Gi) and TETRA-GPRS/3G (Gp) (Gr) interfaces [5].

5.4 Video Encoding Techniques

151

Fig. 9. Integrated TEDS TETRA V+D, GPRS/3G and IP technologies


5.4.1 Background Transmitting video sequences over mobile channels involves several problems since the video data size compared to channel throughput is significantly high. The bandwidth of an original video sequence is too large to fit the bandwidth available in TETRA. Therefore, the need for video compression and coding is significant in wireless telecommunication. Compression renders the data sensitive to any errors and since mobile channels are characteristically noisy, the quality of the received video can be severely degraded. Coding is typically applied to protect the data and allow the decoder to correct any errors that have occurred. However, error correcting codes insert undesirable additional redundancy into the bitstream. Hence, there is a need to design error resilience schemes that keep redundancy to a minimum while offering a high level of visual quality in the decoded video. Methods such as layered coding [26], error concealment [27] and motion vector protection [28] have been developed to address this issue. Compressing the video information will have an effect on the quality of the motion video. The quality of a video sequence is represented by three main factors:

152 -

-

-


Frame Rate. It is the number of pictures being shown per second that give the viewer the impression of motion. The higher frame rate, the smoother playback. The frame rate for full motion movies shown in theaters is 24 frames per second (fps). Color Depth. It describes the number of bits used to represent the color of a single pixel in a bitmapped image or video frame buffer. Color depth is also known as bits per pixel (bpp), particularly when specified along with the number of bits used. Increasing color depth results in higher color quality at the expense of display speed and responsiveness. Resolution of the frame. The number of individual picture elements (pixels) which consist the image dimensions, usually represent the resolution. The pixel resolution is described with the set of two positive integer numbers, where the first number is the number of pixel columns (width) and the second is the number of pixel rows (height), for example as 640x480.

Reducing the frame rate to a full motion movie by half (12fps), will also reduce the size of the video information by half as well. However, a reduction in the frame rate will also influence the smoothness of the playback. A frame rate below 10fps will cause abnormality in the playback. Furthermore, the eye is more sensitive to the luminance (Y-component) of an image than the changes in the color (U (Cb) and V (Cr)). This means that one can reduce the number of bits in representing the color depth without the human eye noticing it. This will additionally reduce the size of the video information. Lastly, diminish the resolution of each frame will compress the video further. For example, scaling down a full screen PC display (640x480) to a resolution of 160x120 will reduce the size of the video information to one-eight of its original size. Beneath it is illustrated the data rate by various frame rates, resolutions and color representations. Table 3. Different compression ratios Compression None 1/12 1/48

Frame Rate 30 fps 15 fps 15 fps

Resolution 640x480 pixels 320x240 pixels 160x120 pixels

Color Depth Data Rate 24 bit 216 Mbit/s 16 bit 18 Mbit/s 8 bit 4.5Mbit/s

The quest to maximize information throughput on given limited channels requires effective and complex compression techniques that achieve higher video quality and higher available capacity at the medium. Further and efficient compression of the video information is achieved by the use of a video codec. A codec is a software component, which translates video


153

between its uncompressed form and the compressed form in which it is stored or transmitted by use of several algorithms. The video codec takes advantage of the fact that the information of each frame in a video sequence varies little compared with the neighborhood frames and with the group of frames that comprises the same scene. Thus, instead of storing the entire frame, it stores only the changes between the current frame and the previous one. This compression method is called “inter-frame” compression (temporal compression) and is also used together with “intra-frame” compressing (spatial compression) techniques such as reduction in the number of bits used to represent the colors in a frame. 5.4.2 Compression standards overview H.263

H.263 is a video codec designed by the ITU-T as a low-bitrate encoding solution for videoconferencing. This codec is an extension of the ITU-T recommendation H.261 and targets video encoding bitrate below 64 Kbit/s. Comparing to H.261; it provides better picture quality at low bit-rates [27] and better error recovery performance [28]. It was first designed to be utilized in H.324 based systems (PSTN and other circuit-switched network videoconferencing and video telephony), but has since found use in H.323 (RTP/IP-based videoconferencing), H.320 (ISDN-based videoconferencing), RTSP (streaming media) and SIP (Internet conferencing) solutions as well. H.263 was developed as an evolutionary improvement based on experience from H.261, the previous ITU-T standard for video compression, and the MPEG-1 and MPEG-2 standards. Its first version was completed in 1995 and provided a suitable replacement for H.261 at all bitrates. It was further enhanced in projects known as H.263v2 (also known as H.263+ or H.263 1998) and H.263v3 (also known as H.263++ or H.263 2000), provide better performance over packet-switched networks and permit scalable bit streams [26]. The next enhanced codec developed by the ITU-T (in partnership with MPEG) after H.263 is the H.264 standard, also known as AVC and MPEG-4 part 10. As H.264 provides a significant improvement in capability beyond H.263, the H.263 standard is now considered primarily a legacy design (although this is a recent development). Most new videoconferencing products now include H.264 as well as H.263 and H.261 capabilities.

154


Video frame structure- Motion estimation and compensation

While H.261 only supports two standardized picture formats (CIF and QCIF), the H.263/H.263+ adopts three additional standardized picture formats (sub-QCIF, 4CIF and 16CIF). Furthermore, it is also possible to negotiate a custom picture format [29]. Each picture is coded based on luminance and two chrominance difference components (Y, Cb and Cr). The luminance (Y) represents brightness information in picture and the chrominance (Cb and Cr) represents color difference components. Each picture in the video sequence is divided into a group of blocks (GOB). For a picture of QCIF resolution, the number of GOB is 9. Each GOB is further divided into macroblocks. Each macroblock consists of four luminance blocks of 8 pixels × 8 lines and the spatially corresponding 8 pixels by 8 lines of Cb and Cr [29]. H.263 supports inter-picture prediction, which is based on motion estimation and compensation. The coding mode in which temporal prediction is applied is called INTER-mode (P-pictures) – where only the difference between original pictures and motion-compensated predicted pictures needs to be encoded. Where no temporal prediction is applied (no reference to any other picture), the coding mode is called INTRA-mode (Ipictures) coding. In the case of the PB-mode, B-pictures are always-coded in INTER-mode. B-pictures are bi-directionally predicted, i.e. predicted both from the previous decoded P-picture and the P-picture currently being decoded.

Fig. 10. Picture prediction

If the areas in the picture do not change (e.g. background) then those parts will not be encoded. Further reduction of the size of the picture information is achieved by attempting to estimate where areas of the previous picture have moved to in the current picture and compensate for this


155

movement. Two-dimensional motion vectors (MV) is used to denote this motion information. A motion estimation (ME) module compares each macroblock in the current picture with its surrounding area in the previous picture to find matching areas and moves these areas into the current macroblock by use of the motion compensator module. Further, the motion compensated macroblock is subtracted from the candidate macroblock. The most widely used method to resolving the best matching blocks in motion estimating is the Sum of Absolute Difference (SAD) due to this method demands low computational complexity and performance: SAD=

1 M ⋅N

M

N

∑∑ X m =1 n =1

m,n

− X m+i ,n + j

(5.1)

Where N, M are the dimensions of the block, Xm, n is the value of the block element located in row m and column n and Xm+i, n+j is the value of the block located in row m+i and column n+j in reference to the preceding picture. To accomplish a better error resilient video coding method, the updated H.263+ introduces the concept of scalable video coding. Scalable video coding makes it possible to separate the video bit stream into multiple logical channels, with the intention that some data can be discarded without impairing the video representation. The types of picture scalability H.263+ employ are temporal, spatial and SNR scalability. Temporal scalability supplies the means for enhancing perceptual quality by increasing the video picture rate. Temporal scalability is achieved using bi-directionally-predicted pictures (B-pictures), which are inserted between anchor picture pairs (between I-pictures and P-pictures) as illustrated in Figure 11. For that reason, for equal quantizing level, this property generally results in improved compression efficiency as compared to that of P-pictures [29], which is only forward predicted.

I

B

P

B

P

......

Fig. 11. Temporal scalability

Moreover, B-pictures are not used as reference pictures for the prediction of any other pictures. This property allows for B-pictures to be

156


discarded if necessary, without impacting the visual picture quality of the future pictures. However, the use of temporal scalability adds structural complexity, larger memory is required and additional delay is introduced. Spatial and SNR scalability are closely related. The distinction is that spatial scalability provides increased spatial resolution, while SNR scalability involves the creation of multiple bit streams. This allows recovery of coding errors or the difference between an original picture and a reconstructed one. This is achieved through the use of a finer quantizer for encoding the enhancement layer to encode the difference. The extra data serves to increase the signal-to-noise ratio of the video picture, and hence, the SNR scalability [29]. Enhancement layer

EI

EP

EP

Base layer

I

P

P

Fig. 12. SNR and spatial scalability

The vertical arrows from the base layer illustrate that the picture in the enhancement layer is predicted from a reconstructed approximation of that picture in the reference (base) layer. If prediction is only formed from the base layer, then the enhancement layer picture is referred to as an EIpicture. It is possible, however, to create a modified bi-directionallypredicted picture using both a prior enhancement layer picture and a temporally simultaneous lower layer reference picture. This type of picture is referred to as an EP-picture or “Enhancement” P-picture [29]. Spatial scalability makes it possible to encode bit streams of more than one resolution to meet varying display requirements/constraints for a wide range of users. Considering a videoconference were the participants use display tools of different resolution capability (e.g. a TETRA terminal of QCIF resolution and a computer screen with CIF resolution capability), a streaming server can send video of different resolution to the participants. As mentioned earlier is Spatial and SNR scalability are essentially the same. The difference is that a spatial enhancement layer tries to recover the coding loss between an upsampled version of the reconstructed base layer picture and a higher resolution version of the original picture [27]. For example, the base layer can be of QCIF resolution and the enhancement layer may be of CIF resolution. The base layer must be scaled in a way that the


157

picture in the enhancement layer can correctly be predicted from it. In H.263+ the resolution can be increased by the factor of two both directions (horizontally and vertically) separately or concurrently. Since there is very little syntactical distinction between pictures using SNR scalability and pictures using spatial scalability, the pictures used for either purpose are called EI- and EP-pictures [29]. Other enhancements, which have an influence on the error resilience of the codec, are listed. • Slice Structured (SS): this mode provides enhanced error resilience capability in order to make the bitstream more amenable for use with an underlying packet transport delivery, and to minimize video delay • Reference Picture Selection (RPS): Improves the performance of realtime video communication over error-prone channel by allowing temporal prediction from pictures other than the most recently-sent reference picture. In error-prone channel environments, this mode allows the encoder to optimize its video encoding for the conditions of the channel. • Independent Segment Decoding (ISD): Allows a picture to be constructed without any data dependencies which cross GOB or multi-GOB video picture segments or slice boundaries. This mode provides error robustness by preventing the propagation of erroneous data across the boundaries of the video picture segment areas. Error resilience and robustness in error prone environments

In environments where the available bandwidth is limited and the transmission suffers from bit errors, the need for post-processing mechanism is present. These post-processing mechanisms are also known as Forward Error Correction (FEC) tools because they provide the means for error correction at the encoder for use in the decoder. Before any attempt to conceal the errors at the decoder, the errors must be detected. The ITU-T H.223 can control errors for various streams. A media packet in H.223 is called an Adaptation Layer Protocol Data Unit (AL-PDU), which is consisted of an optional control field, media payload (which in this case is video information) and a Cyclic Redundancy Check (CRC) checksum. The H.223 is capable of detecting errors or loss of media packets (through block numbering in the video bit stream) using CRC checksum information and pass this information on to the decoder. The decoder can then choose whether to drop the complete packet if errors are detected or indicated [27]. When detection of errors fails in the multiplexing layer, media packets containing errors are passed on to the decoder. The decoder can still detect bit errors using syntactic or semantic violations of the bit stream. These include [30]:

158


Motion vector outside allowable range. Invalid VLC table entry. - Discrete Cosine Transform (DCT) coefficient out of range. - Number of DCT coefficients in a block is exceeded. Techniques to accomplish resynchronization due to errors are also introduced in H.263+. The use of resynchronization enables the decoder to decode valid video information after an error has occurred. Usually the location of decoded bits within the picture cannot be determined because when an error occurs, the number of missing symbols is not known. To overcome this problem, unique resynchronization code words in regular intervals are inserted. In H.263+ these code words are placed in the GOB header to provide an absolute location in the picture. This introduces concerns due to the fact that the size of the GOB’s is varying. This results in that the resynchronization markers within a bit stream are of changing gaps. -

Error concealment

After detecting the error, error concealment is performed. Error concealment is used to hide visual distortion because of residual errors. There are two basic methods to perform error concealment: - Spatial error concealment - Motion compensated temporal error concealment One or both of these approaches can be used depending on the condition. When spatial error concealment is used, the missing pixel values are reconstructed using neighboring GOBs spatial information. In temporal error concealment the lost data is reconstructed from the same GOB in the previous frame [30]. MPEG-2 and MPEG-4 MPEG-2

The MPEG (Motion Pictures Expert Group) committee was formed in 1988 by the hands of Leonardo Chairigloione and Hiroshi Yasuda with the objective of standardizing video and audio for compact discs. MPEG-2 became a standard in 1994 with the object to support a wider collection of applications rather than the compact disk storage as it was established by MPEG-1. MPEG-2 is backward compatible with MPEG-1 and the new coding characteristics that were added and made it possible to improve the functionality and quality at higher bit-rates (up to 40 Mbit/s).


159

Thus, prediction modes were developed to support efficient coding of interlaced video; i.e. two interleaved fields are used to scan out one video frame. In addition, MPEG-2 standard introduced scalable video coding. This allows the division of the continuous bitstream into two or more coded bit streams that represent the video at different resolutions, picture quality or picture rates [31]. The coded video information consists of an ordered set of bit streams, called layers. If there is only one layer, the coded video stream is identified as a non-scalable video bit stream. For two layers or more, the coded video information is called a scalable video bit stream [32]. MPEG-2 uses non-scalable syntax with extra compression tools for interlaced video signals. The algorithm uses block-based motion compensation to reduce the temporal redundancy, which it is used both for casual prediction of the current picture from a previous picture and for noncasual, interpolative prediction from past and future pictures. The prediction error is further compressed using DCT to remove spatial correlation before it is quantized. Finally, an entropy encoder combines the motion vectors with the quantized DCT information and encoded using variable length codes [32]. The scalable syntax is designed to accomplish support for applications beyond those supported with the non-scalable syntax (single layer coding). The objective of this coding method is to offer interoperability among various services and provide additional support for receivers with diverse display capabilities [33]. Applications such as in the area of video telecommunication, internetworking of video standards and video service hierarchies with multiple spatial, temporal and quality resolutions are some the noteworthy application areas [32]. Some receivers are not capable to reconstruct the full resolution video. With the scalable coding syntax, these coders have the possibility to decode subsets of the bitstream to playback the video at lower or temporal resolution or with lower quality. Table 4. Typical parameters for MPEG-2 standard Standardized Main Application Spatial Resolution Temporal Resolution Bit Rate

MPEG-2 1994 Digital TV (and HDTV) TV (4 x TV) 720 x 576 pixels (1440 x 1152 pixels) 50-60 fields/s (100-120 fields/s) 4 Mbit/s

160


Quality Compression Ratio

(20 Mbit/s) Comparable to NTSC/PAL 30-40 (30-40)

MPEG-4

MPEG-4 (together with H.263) has been selected as a multimedia standard for 3rd generation wireless network by Third Generation Partner Project (3GPP). MPEG-4 is an ISO/IEC standard developed by MPEG in 1998. This standard was made with the purpose to be the next standard in the world of multimedia, which supports very low bit-rates.. This standard supplies the users with the opportunity to achieve a variety of forms of interactivity with audio and video information from a scene and to merge synthetic and natural audio and video content [34]. MPEG-4 is designed to provide high degree of flexibility and extensibility in order to take the advantage of the rapidly evolving technologies such as in wireless communication systems. Dissimilar with MPEG-2, which focused on better compression efficiency, MPEG-4 provides an efficient video coding covering the large from the very low bit rates of wireless communications to bit rates and quality levels beyond high definition television (HDTV). MPEG-4 is targeted to be 5-64 kbit/s for mobile applications and up to 4 Mbit/s for TV applications. The standard introduces new video functionalities which can be grouped in to three classes: content-based interactivity, coding efficiency and universal access. • Content-based interactivity MPEG-4 makes it possible to access and manipulate visual-objects in a compressed domain rather than decoding the video sequence to the original format before encoding in to desired compressed bitstream as it was in MPEG-2. This results in a reduction in complexity, reduced bandwidth requirements, low delay, no quality loss and object-based manipulation [35].Further, MPEG-4 provides the means for harmonious interaction of video objects of various origins, which can be either, natural or synthetic [36]. This means that it is possible to encode a video sequence in a way that allows separate decoding and reconstruction of objects in a scene and manipulation of the original scene by simple operations on the bit stream. This is possible because the bit stream is “object layered” and the shape and transparency of each object are described in the stream of each object layer. Therefore, the receiver of a bit stream has the ability to choose if the bit stream shall be reconstructed fully or in a manipulated form. Finally, in the terms of content-based interactivity, MPEG-4 provides efficient


161

methods to randomly access, within a limited time and with fine resolution, parts (e.g. frames or objects) from a video sequence [35]. • Coding efficiency In networks where the available bandwidth is rather limited (wireless networks), it is important that improved coding efficiency is provided. MPEG-4 offers tools and algorithms that offer better compression without affecting the quality of the video. This makes it possible for use of applications such as video conferencing and streaming in mobile networks. In, addition MPEG-4 provides methods to code multiple views of a scene efficiently. • Universal access MPEG-4 provides error robustness capabilities for storage- (e.g. videoon-demand applications) and communication-applications (e.g. video conferencing applications) in heterogeneous error-prone environments that may suffer from severe error conditions (e.g. long error bursts cause by to noise, fading, shadowing and interference). Structure of the video coding algorithm

An MPEG-4 Verification Model (VM) was defined in order to evolve video-coding techniques in a combined effort. The MPEG-4 verification model specifies the input and output formats for the non coded information and the format of the bit stream containing the coded information. Furthermore, it demonstrates a defined core video coding algorithm platform for the standard in progress. The MPEG-4 video verification model’s main features include: -

-

-

-

Standard Y:U:V luminance and chrominance intensity representation of regularity sampled pixels in 4:2:0 format. The intensity of each Y, U or V pixel is quantized into 8 bits. The size and shape of the image depends on the application. Coding of multiple “Video Object Planes” (VOP’s) as images of arbitrary shape to support many of the content based functionalities. Coding of shape and transparency information of each VOP by coding binary or grey scale alpha plane image sequences using a particularly optimised Modified Reed Code method. Support for intra (I) coded VOP’s as well as temporally predicted (P) and bi-directionally (B) predicted VOP’s. Standard MPEG and H.263 I, P, B frames are supported as special case. Support of fixed and variable frame rates of the input VOP sequences of arbitrary or rectangular shape.

162 -

-

-

-


8x8 pixel block-based and 16x16 pixel Macroblock-based motion estimation and compensation of the pixel values within VOP’s, including provisions for block overlapping motion compensation. Texture coding in I, P and B VOP’s using an 8x8 DCT or alternatively a shape adaptive DCT (SADCT) adopted to regions for arbitrary shape, followed by MPEG-2 and H.263 like quantization and run-length coding. Efficient prediction of DC and AC coefficients of the DCT in intra coded VOP’s. Support for efficient static as well as dynamic SPRITE prediction of global motion from a VOP panoramic memory using 8 global motion parameters. Temporal and spatial scalability for arbitrary shaped VOP’s. Adaptive macroblock slices as well as improved bit stuffing and motion maskers for resynchronization in error prone environments.

Video Object, Video Object Layer (VOL) and Video Object Plane (VOP)

In contrast to the frame-based video coding algorithms of MPEG-2 and H.263, MPEG-4 is an object-based algorithm. The MPEG-4 video-coding algorithm provides an efficient representation of visual objects of arbitrary shape, with the objective to support content-based functionalities. The content-based functionalities make it possible to separate the encoding and decoding of physical objects in a scene (i.e. each physical object is encoded/decoded individually). The representation architecture envisioned in the verification model is based on the concept of Video Object Planes (VOP’s). VOP’s is a number of arbitrary shaped image regions where each region can cover particular video content, i.e. describing physical objects or content within scenes. The video input can be a VOP image region of arbitrary shape where the shape and the location vary over time. For each VOP the encoder processes the shape, motion, and texture characteristics. The shape information is encoded by bounding the VO with a rectangular box and then dividing the bounding box into Macro Blocks (MBs). Each MB is classified as lying inside the object, on the object’s border, or outside the object (but inside the bounding box). The texture coding is done on a per-block basis similar to the “frame-based” standards, such as H.263. In an Intracoded (I) VOP the absolute texture values in each MB are DCT coded. The DCT coefficients are then quantized and variable-length-coded. In forward Predicted (P) VOPs each MB is predicted from the closest match in the preceding I (or P) VOP using motion vectors. In Bi-directionally predicted


163

(B) VOPs each MB is predicted from the preceding I (or P) VOP and the succeeding P (or I) VOP. The prediction errors are DCT coded, quantized, and variable-length-coded. The I, P, and B VOPs are arranged in a periodic pattern referred to as Group of Pictures (GoP). A typical GoP structure is IBBPBBPBBPBB. For the transmission the shape, motion, and texture information is multiplexed at the MB level, i.e. for a given MB the shape information is transmitted first, then the motion information, and then the texture information, then the shape information of the next MB, and so on. VOP’s that belongs to, or describes the same physical object in a scene is, commonly known as Video Objects, VO’s. Each VO may have two or more scalability layers, where the lowest level is a base layer and the other layers function as enhancement layers, referred to as Video Object Layers (VOL’s). The VOL’s contains the shape, motion and texture information of VOP’s that belongs to the same VO. Additional, the VOL’s contains relevant information that is necessary to identify each of the VOL’s and how the different VOL’s are organized at the receiver to recreate the entire original sequence. This makes it possible to exploit separate decoding of the VOP’s and allows content-based manipulation without the need for transcoding.

Fig. 13. MPEG GOP Structure

164


Fig. 14. Coding of video sequence using MPEG-4

In order to support separate decoding of VO’s the data associated to the shape, motion and texture for the VOP’s (belonging to the same VO) are coded into separate VOL-layers. The coding scheme MPEG-4 provides if the input video sequence only consists of rectangular sized images have the same structure comparable to MPEG-1 and MPEG-2 video coding algorithms. In this case, the shape information for the VOP’s is not transmitted to the decoder.


165

Fig. 15. MPEG-4 macroblock grid for a VOP image

The shape information is presented in two formats [38]: - Binary format: each pixel has value information whether the pixel is within the VO or not. - Grey scale format: the pixel value can have a value ranging from 0255. Value 0 indicates transparency and 255 indicate that the pixel is opaque. Values between 0 and 255 indicate to an intermediate intensity of transparency. The shape information is encoded by bounding the VOP with a rectangular box (reference window) and then dividing the bounding box into shape blocks. Motion estimation and compensation

MPEG-4 gains benefit of block-based motion estimation and compensation techniques to explore temporal redundancies of the video information in the separate VOP layers. The principle for motion estimation and compensation is to find the best-matched block in the previous VOP frame for every block in the current one. A motion vector is estimated for each current block and indicates the displacement of the previous block. The motion compensated prediction error is estimated by subtracting the pixel

166


values of the predicted block from the pixel values of the current block [37].

Fig. 16. Motion Estimation and Compensation

Motion estimation and compensation in MPEG-4 introduce furthermore a “padding technique” that excludes the pixels not belonging to the active VOP area. The padding procedure can be regard as a prediction of the pixels outside of the VOP based on pixels inside it. Texture coding

Texture coding is performed on the prediction error block. The coded motion vector and the coded texture information are then sent to the decoder. The decoder can then make use of to reconstruct the current block by adding the quantized error prediction block to the predicted block according to the motion vector. Scalable coding of video objects

Content-based scalability allows end-users to select from various bandwidths, display capabilities or display requests to allow video database browsing and multi-resolution playback of the video content-based on the bandwidth resources available in the particular mobile network. In addition, object-based functionalities allow coding of VO’s of arbitrary shapes.


167

The scalability framework is referred to as generalised scalability and includes the spatial and temporal scalabilities. • Spatial scalability Two layers are provided where the “mid processor” performs spatial up or down sampling of the input. Each of the layers supports a VOP at different spatial resolution scales. The downscaled version of the VOP is encoded into the base layer of the bit stream and is encoded with a non-scalable coding technique. The VOP in the enhancement layer is either encoded as a P-VOP (inter predicted-VOP) or a B-VOP (bi-directionally predicted VOP).This makes it possible for an end-user to choose to display VOP signals with full resolution or VOP signals that are reconstructed by decoding the base layer bit streams. • Temporal scalability The temporal scalability offers scalability of the temporal resolution and figure 26 illustrates the objective behind the temporal scalability scheme. In this case the “mid processor does not perform and spatial resolution conversion [36]. Layering is achieved through temporal prediction for the enhancement layer based on coded video from the lower layers. Using MPEG-4 temporal scalability makes it possible to supply different display rates on different VOL’s within the same video sequence.

Fig. 17. Spatial scalability scheme

168


Fig. 18. Temporal scalability scheme

Error resilience and robustness in error prone environments

The error resilience tools provided in MPEG-4 are divided into three different areas: resynchronization, data recovery and error concealment. • Resynchronization Resynchronization is necessary to be executed after errors have been detected. The resynchronization approach implemented in MPEG-4 is a packet approach and is based on providing periodic resynchronization markers at every N bits (N is an arbitrary integer value) throughout the bit stream. The use of resynchronization markers allows the decoder to move forward to the next resynchronization marker when it detects an uncorrectable error, thus reducing the amount of data that must be discarded when a resynchronization loss occurs. When the video frame is divided into video packets with resynchronization markers in-between, less data will be lost due to the fact that the video packets can be decoded independently [34].

Fig. 19. Normal video frame. A lot of data lost in the frame


169

Fig. 20. Resynchronization markers within a video frame divided into video packets

• Data recovery Data recovery mechanism attempts to recover data that in general would be lost. This mechanism does not provide error correction, but techniques that can encode the data in an error resilient behavior [38]. A tool which provides it in MPEG-4 codec is the Reversible Variable Length Coding (RVLC) where the variable length code-words are designed so they can be read in both forward and reverse manner. In transmissions where an error has corrupted a section of the video packet, all the data between two resynchronization markers will normally be lost without use of RVLC. Though this will improve the error resilience significantly, the compression efficiency will be reduced.

Fig. 21. Example on use of RVLC

• Error concealment An extremely vital element of any error robust video codec is error concealment. The error concealment tool is dependent on the performance of the resynchronization scheme and if the resynchronization method is able to effectively identify the error then the error concealment problem becomes a great deal more tractable [36]. To further enhance the error

170


concealment capabilities and advance the ability of the decoder to identify errors an approach called “data-partitioning” is exploited. Data partitioning separates motion and header data from texture data within each of the video packets by inserting a second resynchronization marker. Data partitioning is signaled to the decoder in the VOL. If an error occurs and the texture information is lost, data partitioning makes use of the motion data to conceal the error by using the motion information to motion compensate the previously decoded VOP [36].

Fig. 22. Data partitioning

5.4.3 Encrypted Video over TETRA The TETRA system uses end-to-end encryption in addition to the air interface encryption to provide enhanced security. End-to-end encrypted continuous data, such as video, requires synchronization of the key stream at the receiver to the incoming encrypted data stream from the transmitter. Apart from the video coding synchronization mechanisms (e.g. MPEG-4, H.263), the TETRA system uses a synchronization technique known as frame stealing to provide synchronization to end-to-end encrypted data. Nevertheless, the frame stealing process degrades the quality of video and is not suitable for transmission of secure video. In order to overcome frame stealing technique’s drawbacks, alternative synchronization and recovering techniques have been proposed which are more suitable for video data like frame insertion and fly wheeling [39]. Communications over wireless channels are very noisy and time varying resulting in errors in the received data and as a result loss of synchronization to the incoming data stream. Although TETRA speech codec is capable of tolerating some loss of data without important effect on the comprehension when a frame stealing technique is used, this strategy is not suitable for video. Each video frame contains information about the differences between consecutive frames, thus losing part of such a frame results in incorrectly interpreting incoming video frames. Moreover, video encoding


171

procedure is taken place at the application layer unlike speech which makes difficult the use of frame stealing technique as it is done in the data link layer. That is why frame insertion technique is introduced for providing synchronization to encrypted video streams. Also, fly wheeling technique can be used to recover synchronization from dropped packets. Frame Insertion Technique

In frame insertion technique synchronization frames are inserted between successive video frames before transmission. At the transmitter side, the encryption block receives the video frames, as its input, and the encrypted data are multiplexed with the synchronization data according to control signals from the appropriate block. The multiplexer waits for the current encrypted frame to finish routing before it routes the synchronization frame. The synchronization frame includes the frame marker, the current key number, the algorithm number, the current IV and the checksum which is obtained from the previous four fields. On the other side, at the receiver the first step is the check of the received frame type, video or synchronization frame. This is done by verifying the received frame size, the frame market and the checksum. When a synchronization frame is arrived the current key and the algorithm are changed according to the received key number and the algorithm number, while the IV of the received decryption unit is updated according to the received IV. Figures 23a and 23b describe the above mechanisms.

Fig. 23. a. Synchronization mechanism at transmitter. b. Synchronization mechanism at receiver

Fly Wheeling Technique

At the case of dropped packets, the main problem is to determine the size a dropped packet when the packets are of variable length. This is necessary

172


to correctly update the IV table of the decryption unit in order to correctly decrypt the successive packets after the dropped packet. The solution to this problem is to generate fixed length key stream segments (KS 1, KS2 ...) to encrypt the variable length data packets (DP1, DP2 ...) Figure 24. The lengths of each of these key stream segments is independent of the data packet sizes and should be equal to an allowable or expected maximum data packet sue. After encrypting each data packet whose sue is less than the expected maximum, the excess portion of the key stream segment is discarded. In a practical implementation this can be achieved as following. After each data packet is encrypted, the IV is updated ‘n’ times. In a block cipher based system, n is equal to the difference between a predetermined maximum data packet size and the current data packet size in terms of the block size of the cipher. “n” is the number of rounds that block cipher should operate. In this case, one round of encryption is equivalent to running the block cipher once. For example, if the maximum packet size is 20000 bytes and the block size of the cipher is 8 bytes and the current data packet size is 16000 bytes, then; n = (max packet size - current packet size)/ block size = 500 = (20000 - 16000)/8. In a stream cipher based system which produces one bit of key stream per round of operation of the cipher, n is equal to the difference between a predetermined maximum data packet size and the current data packet size in terms of bits per packet. Here, one round is equivalent to the operation involved in producing one bit of key stream from the key stream generator, i.e. one round equals one bit. For example, if the maximum packet length is 25000 bits and the current data packet size is 15000 bits, then; n = max packet size - current packet size =25000-15000 = 10000 These two processes effectively generate constant length key stream segments to encrypt each variable length data packet. The length of each key stream segment is equal to the maximum allowable or expected data packet size. In this case, the starting IV value of encryption unit to encrypt a particular data packet is independent of the size of the previous data packet. Therefore, the receiver can independently determine the starting IV value of each received packet if it can detect dropped packet. This is facilitated by assigning a sequence number to each encrypted data packet. The sequence number counter is started from zero and incremented by one for each data packet transmitted up to a predetermined maximum value. At the transmitter, and receiver, the sequence number counters are reset to zero when they reach the maximum value. A data packet received out of sequence indicates a dropped packet. Therefore, if the packet N+1 is received immediately after the packet N-1 (i.e. packet N is dropped), the key stream segment KSN, which is used to encrypt the packet N will be


173

discarded Figure 25. Similarly, if the packet N+2 is received immediately after the packet N-1 (i.e. packets N and N+1 are dropped), then the key stream segments KSN and KSN+1 should be discarded. Here, if an encrypted data packet is received out of sequence, the IV of the decryption unit should be updated accordingly before decrypting the particular packet. For example, consider a scenario where a block cipher with block size equal to 8 bytes is used in a system with a maximum data packet size of 20000 bytes. If the packet N+1 is received immediately after the packet N-1, then the IV is updated 2500 {=20000/8) times before decrypting the packet N+1. Similarly, if the packet N+2 is received immediately after the packet N-1, then IV should be updated 5000 {= (2×20000) / 8} before decrypting the packet N+2. The fly-wheeling mechanism can be combined with the sequence numbering scheme in such away that enables the receiver to compute the starting IV of a received packet from its sequence number. In this case, the transmitter does not have to send the IV to the receiver in order to resynchronize it as the sequence numbering scheme can compensate for dropped packets. The only requirement in this case is the initial synchronization. After initial synchronization end-to-end encryption synchronization can be achieved without insertion or stealing. Initially a random IV (IVR) should be negotiated between the transmitter and the receiver. This can be done either by securely sending the IVR to the receiver by encrypting it, or computing it by combining a random seed with the session key (for example by hashing). The initially negotiated IVR should be the starting IV of the first packet. After that fly-wheeling can be employed to determine the starting IV of the subsequent packets. This scheme can be implemented using a block cipher operating in counter mode where instead of the feed back a counter is used as the input of the encryption process. To Fly-wheel the counter should be incremented a predetermined (M) times for each packet. Here also the starting IV for the first packet is IVR. The staring 1V for the subsequent packets can be calculated by multiplying the packet number N by M and adding it to IVR. Therefore, the starting IV of packet N denoted by IVNS in the series of packets 0,1,2,3, ....., N- 1 , N, N+1, ... can be calculated as follows; IVSN = IVR + N×M The value IVSN, can be calculated separately and can be loaded to the counter, so that decryption unit does not have to operate M times for each packet, thus saving some processing. Also, the packet numbers can be nonsequential. They can vary randomly, and a random number generator can be used to assign numbers to the packets. In this case N will be the random number assigned to each packet. Still the above equation for IVSK is valid.

174


Since the packet number field (e.g. 16 bits) will not be as large as the IV field (e.g. 64 bits), a new IVR should be assigned when the packet count reaches its maximum (i.e. when it wraps around). This can be done by either the transmitter generating and sending a new IVR to the receiver when the packet count wraps around or deriving a new IVR from the old IVR, (for example by hashing the old IVR to obtain a new IVR).

Fig. 24. Key generation for data packets

Fig. 25. Desynchronizing after loss of data packets at the receiver

5.5 Performance Analysis of video broadcasting over TETRA TETRA already provides voice and data transmission (where the data consists of photos, fax, vehicle locations, maps etc.). The incorporation of TEDS standards into current systems will provide quality real-time mobile video transmission, which will correspond to a third generation of PMR systems and will offer several advantages to PMR users as it has been already discussed. A common example that is included in the literature is the case of “Technocop”, a police officer who wears a helmet with a visor that

5.5 Performance Analysis of video broadcasting over TETRA

175

can receive video and other types of data like maps, locations and photos [22]. In addition, there would be a camera embedded in the helmet so he could send video sequences of the crime scene to other officers. Another example of the importance of real time video transmission for PSDR agencies is during an accident where a camera could be used to transmit a video of the accident scene back to the police and ambulance services which could then accurately assess the situation and dispatch the relevant medical equipment and aid [18],[19]. However, there are several problems involved in transmitting video sequences over mobile channels. Since the video data is so voluminous, it must first be compressed before transmission. This compression step renders the data sensitive to any errors and since mobile channels are characteristically noisy, the quality of the received video can be severely degraded. Channel coding is typically applied to protect the data and allow the decoder to correct any errors that have occurred. However, there is limited bandwidth available in mobile networks and error correcting codes insert undesirable additional redundancy into the bitstream. In this section we provide a performance analysis of video transmission over TEDS network. In particularly, based on a real video source we analytically estimate the packet loss that each user suffers, when the traffic load increases. The perceived QoS assessment based on specific network condition and traffic loads provides a performance indication of video traffic over TETRA [25]. 5.5.1 Performance Evaluation The video traffic from a single MPEG-4 video source is represented by a Markov statistical model [42],[43]. In order to evaluate the performance of the video traffic in an TEDS system, several homogeneous and mutually independent statistical video sources have been used. Thus, the aggregate video traffic model can be obtained by creating multiple instances of the single traffic model. The multiple traffic model of the network can be represented as a Markov chain. TEDS system is considered to be ideal and we assume Rayleigh fading multipath and White Additive Gaussian noise over the wireless medium. The following Figure 26 shows the multiplexing model considered during the analysis [43].

176


Fig. 26. N video sources multiplexed to a single stream

An MPEG-4 open source video tool has been used to encode three different QCIF video sequences (176x144) (highway, mother-and-daughter, news). Each video sequence is encoded as VBR using the MPEG-4 codec at 64 kbps and q-step size equals to 8. The video frames are sent every 33ms for 30fps video. The real video data was considered as a circular list and each video source uses a different starting point such that the generated traffic from each source was not identical a packet-by-packet basis so that there is low correlation among the video sources. The simulator has been running several times from a different starting time for each video source, so that the collected statistical data do not depend from the correlation of the generated data. In the following Figure 27 the basic simulation components and parameters are illustrated.


177

Fig. 27. Simulation components for MPEG-4 coded video transmission simulation.

The total simulation is long enough to capture long-range dependence effects of the aggregate video traffic. Since the focus is on evaluating the performance of actual video traffic and the effect of video scheduling, we have selected a random alignment of multiplexed streams rather than the “optimal” frame arrangement among the sources that ensures minimum required bandwidth. Hence, our model is subject to burst errors and bottleneck effects due to congestion as in a real time randomly transmitted video traffic. Figure 28 illustrates the overall packet loss according to the above simulation scenarios. The packet loss has been obtained from real data under 1Mbps TEDS network data rate. The analysis successfully illustrates the effect that the increased traffic load due to new users has on the video packets loss probability. Additionally, in the same figure it can be seen that depending on the context of the video sequence (static images “bridge” or motion sequences “highway”) the effect of network traffic load differs. Larger I-frames in the first case means that if a packet loss occurs in an

178


I-frame type then the loss will be large and it will propagate to the rest frames in the GOP sequence resulting in higher packet loss overall.

Fig. 28. Overall packet loss for TEDS data rate of 1Mbps for traffic load ranging from 5 to 20 video sources. Two video sequences are included (“highway” and “bridge”)

5.5.3 Video Quality Measurements The encoded video bit stream has been exposed to packet loss. The packet loss is determined from the TEDS performance of Figure 28. For the shake of simplicity all sources experience packet loss proportional to the generated bit rates. Standard concealment methods provided by the MPEG4IP open source tool have been used in order to substitute the lost information of the damaged area in each erroneous frame. The video sequence with real data has been continuously monitored in order to track the lost packets per frame. Therefore, at the receiver, the decoding procedure for each frame comprises three states. Our analysis is different from the typical Gilbert model that has been used to simulate channel model in order to evaluate the perceived image quality. In our analysis, the following cases have been considered to estimate the impact of packet losses per frame:


179

• No packet loss and as an effect, no distortions are introduced at the decoded image frame at the receiver. • A single packet loss. This means that a single packet loss occurs within the frame. This packet loss may correspond to certain bits of either lost or corrupted MBs. The number of affected MBs can be determined by calculating the average number of generated bits/MB for each frame type. Without loss of generality, we assume that the position of lost information (start of loss within the frame and end of loss within the frame) follow a uniform distribution. • Multiple packet loss. Several lost packets (more than one packet is lost in each frame) occur consecutively. This results to the loss of several MBs within the frame or even the loss of an entire frame. Such a loss may severely damage the perceived video quality in case of intraframes. In Figure 29 the PSNR lag plot versus frame is illustrated for the “Highway” sequence. The PSNR lag plot of a single real video data source is calculated based on the overall packet losses occurred by all the video sources. Finally, we provide visual results that indicate the degradation of video quality when packet loss occurred.

Fig. 29. Perceived QoS in terms of PSNR for 5 and 20 real video sources transmitting the “highway” sequence in a TEDS 1Mbps data rate

180


Fig. 30. Visual results for “highway” sequence for 5 and 20 real video asources

5.6 Vision for Future Public Safety Communication Systems 181

5.6 Vision for Future Public Safety Communication Systems

5.6.1 Future Trends There is no industry consensus on what future networks will look like but, as far as the next-generation networks are concerned [41], ideas and concepts include: • Transition to an “All-IP” network infrastructure. • Support of heterogeneous access technologies (e.g. UTRAN, TEDS, WLANs, WiMAX, xDSL, etc). • VoIP substitution of the pure voice circuit switching. • Seamless handovers across both homogeneous and heterogeneous wireless technologies. • Mobility, nomadicity and QoS support on or above IP layer. • Need to provide triple-play services creating a service bundle of unifying video, voice and Internet. • Home networks are opening new doors to the telecommunication sector and network providers. • Unified control architecture to manage application and services. • Convergence among network and services. Two important factors have been considered to satisfy all these requirements. The first one regards the interworking of existing and emerging access network under the umbrella of a unified IP-based core network and unified control architecture supporting multimedia services. A proposed solution towards this direction is the Unlicensed Mobile Access (UMA), allowing heterogeneous wireless technologies to interconnect to a core network through a network controller. The second requirement regards IP multimedia subsystem (IMS) evolution in order to cope with requirements imposed by NGN architecture. The initial release of 3GPP IMS was developed only for mobile networks. The increasing demand of interworking between different access devices and technologies led to subsequent releases that defined IMS as core independent element and a key enabler for applying Fixed Mobile Convergence (FMC). FMC comprises of two attributes: using one number, voice/mail and seamless handover of multimedia sessions. In the B3G/4G vision, IMS is required to become the common architecture for both fixed and mobile services. Towards this end the ETSI Telecoms and Internet converged Services and Protocols for

182


Advanced Networks (TISPAN) is also producing new functionality extensions for the IMS (ETSI TISPAN). 5.6.2 All-IP convergence The major benefits that drive towards an all-IP based core network are the following: • Cost saving on ownership and management: network operators need to own and manage one single network, instead of multiple. • Cost saving on transport e.g. the cost to provide IP transport is lower. • Future proof: it can be claimed that the future of backbone network, both for voice and data, is IP based. An IP-based network allows smooth interworking with an IP backbone and efficient usage of network resources. • Smooth integration of heterogeneous wireless access technologies • The IP Multimedia domain can support different access technologies and greatly assist towards fix/mobile convergence. • Capacity increase: the capacity enhancement of IP based transport network is quicker and cheaper. The same is also true to service capacity, thanks to the distributed nature of the service architecture. • Rich services: the benefits of VoIP are available for improved and new services e.g. voice/multimedia calls can be integrated with other services, providing a powerful and flexible platform for service creation. • Enable peer-to-peer networking and service model. This hybrid network architecture would allow the user to benefit from the high throughput IP-connectivity in ‘hotspots’ and to attain service roaming across heterogeneous radio access technologies such as IEEE 802.11, HiperLan/2, UTRAN and GERAN. The IP based infrastructure emerges as a key part of next-generation mobile systems since it allows the efficient and cost-effective interworking between the overlay networks for seamless provisioning of current and future applications and services. Furthermore, IP performs as an adhesive, which provides global connectivity, mobility among networks and a common platform for service provisioning across different types of access networks. The development of an all-IP interworking architecture, also referred to as fourth-generation (4G) mobile data network, requires specification and analysis of many technical challenges and functions, including seamless mobility and vertical handovers between WLAN and 3G and TETRA technologies, security, authentication and subscriber administration, consolidated accounting and billing, QoS and service provisioning [40],[44].


5.6.3 TETRA – TEDS interoperability An important issue concerning the future public safety networks is to ensure TEDS-TETRA interoperability [12]. However, as it will take time to develop an interoperable, TEDS-capable infrastructure and terminals market, an early decision is essential to maintain the interoperability already achieved in today’s TETRA networks. In order to achieve that TEDS systems should preferably operate in the same frequency band as TETRA voice to enable cost-effective, combined radio terminal implementations that accommodate both voice and high-speed data. Furthermore, the existing TETRA base stations should be upgraded in order to improve the coverage range and extend it to vehicle or vessel mounted and handheld terminals, for voice and for data. The TETRA radio transceivers must also be upgraded in order to run current TETRA voice and data radio interfaces as well as the TEDS radio interfaces and to be able to satisfy the higher capacity needs of TEDS. The key feature to the above technological advances that will enable the full capabilities of TEDS, is that they should be performed seamlessly and with minimum cost. The advantage of using TEDS technology as far as interoperability with TETRA networks is concerned is the fact that no new network elements are required, so service and maintenance costs are not increased. The interface from the TETRA infrastructure to service nodes also does not change. TEDS supports all the data services used in the existing TETRA network. The network is already provisioned to all high-speed users as well and no new provisioning, billing and service management interfaces are required. This is a major saving compared to the alternative of building an overlay high-speed data network for public safety users. Economical upgrade to TEDS highspeed data capability will ensure that TETRA remains viable – in fact, vital – in professional mobile radio communications for years to come. Taking advantage of developments in mobile data services, TETRA users will have a growth path offering continual access to the leadingedge technical solutions. 5.6.4 TETRA over IP As it has been said above IP enables convergence of fixed and mobile, voice and data communications. TETRA standard and even more TEDS allow for convergence over the IP backbone platform [3], [9]. The connectivity to IP results in excellent service quality, system adaptability, harmonised standards, various integrated services, and a new service in the

184


“push-to-talk” mode of communication. Figure 31 below shows a typical topology of TETRA over IP network, where all devices are interconnected via an IP network.

Fig. 31. Typical topology of TETRA over IP network

5.6.5 Integrated TETRA-WLAN system In addition to the proposed integration of TEDS and other cellular systems, it is interesting to study the perspective of incorporating WLAN to TETRA networks, which is deemed as an interesting perspective capable of providing advanced new features to the PSDR communication systems. The ongoing intense WLAN standardization and R&D activities worldwide, which target bit rates higher than 100 Mb/s, enhanced security, improved mobility management, quality of service support, and interworking with cellular networks, justify the fact that WLAN technology will play a key role in the wireless market. Therefore WLAN and TETRA networks integration proposed in [24] may result in a number of benefits that are not possible to achieve otherwise. The proposed system configuration is illustrated in the schematic diagram of Figure 32. The system enables TETRA terminals to interface with the TETRA SwMI over a WLAN radio interfaceas opposed to the conventional TETRA V+D radio interface [2] and over an IP access network [3]. These TETRA-WLAN terminals can employ all typical TETRA services, but can also support a range of brand new services and there is no further


requirements for SwMI in order to support these terminals. Although, conventional TETRA terminals and TETRA-WLAN terminals are alike as far as the network is concerned, the characteristics of the WLAN radio interface enable the later terminals to enjoy extended capabilities and new features. These include high-speed data services, inexpensive and simple provision of simultaneous voice and data services, improved voice quality, better call performance (reduced setup and voice transmission delays), simultaneous reception of many group calls, and so forth [10].

Fig. 32. TETRA-WLAN integration system overview

The key component of the integrated system is the Interworking Function (IWF), which interfaces to the SwMI similarly to a TETRA BS. The IWF also interfaces with one or more WLAN Access Gateways (WAGs) and uses IP multicasting technology to transfer control packet data units (PDUs) and voice packets to the TETRA-WLAN terminals. All terminals can access the typical services provided by the TETRA SwMI by means of a WLAN network interface and the appropriate software drivers and applications. It is important to note that interfacing IWF with a TETRA SwMI that is already built on IP multicast and Voice-over-IP (VoIP) technologies is relatively straightforward [24]. The proposed integration of WLAN and TETRA systems several beneficial feature can be attained. These heterogeneous networks support higher levels of availability, reliability, and network survivability. Specifically, by using dual-mode terminals and two different radio access technologies with overlapping coverage in strategic locations, the service

186


availability is considerably improved in these locations. Also, the deployment of WLAN as an overlay radio access network leads to increased overall system survivability, which is of paramount importance in public safety and disaster recovery communication systems [12]. Increased reliability is also doable, especially in indoor areas where the WLAN can provide better RF coverage and thus ensure less radio link failures and less dropped or missed calls. Second, a variety of revolutionary new features and capabilities can be enabled. In particular, public safety and other mission-critical users will be able to use their dual-mode TETRA-WLAN terminals in the field but also in the office environment, over their private WLAN. Moreover better indoors coverage, increased capacity and novel new services in a costeffective way can be achieved as TETRA services will be provided over the WLAN access. Further, in WLAN hot spots, advance services, such as enhanced voice encoding schemes for better voice QoS, broadband data services, true simultaneous voice and data services, IP applications could be readily exploited and by means of IP multicast, many group calls could be monitored simultaneously by a single radio user. Furthermore, the large air interface capacity of WLAN is able to support many simultaneous TEDS voice and data calls, both efficiently and reliably. The call setup delays and voice transmission delays could be considerably reduced due to the increased bit rates of WLAN. What is more, WLAN networks can easily be integrated with an IP-based network, which is envisioned to form the backbone of the next-generation public safety communication systems for meeting multiple different service requirements. The user will be able to access the TETRA-WLAN hybrid network and the services and applications supported through a large collection of WLAN enabled subscriber devices, such as PCs, PDAs, or even cellular devices with WLAN capabilities. It is well known that dual WLANCellular end terminal devices are quickly emerging in the market and these devices will be able to support TETRA and TEDS services with only limited software enhancements.

5.7 Conclusions In conclusion in this article we presented the important parameters of TEDS system that allow for real time video transmission and through a simulation model we show how the perceived QoS will behave under different traffic loads. The article produced objective QoS assessment in

5.7 Conclusions 187

terms of PSNR versus frame number and visual perceived QoS results for multiplexed video traffic. Analytically, we identified the main reasons why public safety networks require advance capabilities such as multimedia and video transmission support and hence how the enhancements on existing TETRA standards can be proved to be the solution for market demands. We provided a compact analysis of the TETRA and TEDS technical characteristics, advantages and disadvantages that make TEDS the best solution for public safety networks with good coverage, security, availability, scalability, interoperability and QoS support for video transmission. The article summons all the main video compression methodologies that are currently available. Special interest has been shown on MPEG-4 coding scheme, which provides high level compression with the minimum QoS degradation at high speed transmission rates. Thus it makes it ideal candidate for video transmission over the TEDS air interface. As every compression method introduces errors we provided an overview of the most common and well studied error concealment approaches. Finally, we included the most recent and future research scenarios and trends of TETRA-WLAN and TETRA-IP interworking that will further enhance the QoS of video and high data rates of TETRA and TEDS especially.

188


References 1. ETSI EN 300 396 (all parts): “Terrestrial Trunked Radio (TETRA), Technical requirements for Direct Mode Operation (DMO)”. 2. ETSI TS 100 392-2 (V3.1.1): “Terrestrial Trunked Radio (TETRA), Voice plus Data (V+D), Part 2: Air Interface (AI)”. 3. ETSI TS 101 747: “Terrestrial Trunked Radio (TETRA), Voice plus Data (V+D), IP Interworking (IPI)”. 4. ETSI TR 102 021-2: “Terrestrial Trunked Radio (TETRA); User Requirement Specification TETRA Release 2; Part 2: High Speed Data”. 5. ETSI TR 102 491 (V1.2.1): “TETRA Enhanced Data Service (TEDS); Electromagnetic compatibility and Radio spectrum Matters (ERM)”. 6. Brian Oliver, “Standardization for Emergency Communications ETSI Project TETRA”, EMTEL Workshop, February 2002. 7. “Improving TETRA base station coverage with revolutionary radio access solution”, NOKIA white paper, 2004. 8. P R Tattersall, “Professional mobile radio — the BT Airwave public safety service and the path for technology and service evolution”, BT Technol. J. vol. 19, no. 1, January 2001. 9. Borut Dolanc and Matjai Judei, “Professional Mobile system - TETRA over IP and IP over TETRA”, IEEE EUROCON, Ljubljana, Slovenia, 2003. 10. E Lammerts, CH Slump, KA Verweij, “Realization of a Mobile Data Application in TETRA”, Proceedings of ProRISC, 1999. 11. Le Bodic, G. Irvine, J. Dunlop, J., “Resource cost and QoS achievement in a contract-based resource manager for mobile communications systems”, EUROCOMM 2000. Information Systems for Enhanced Public Safety and Security. IEEE/AFCEA, pp. 392-397, Munich, Germany, 2000. 12. Risto Toikkanen, “TETRA - radios for harmonised disaster communications”, 2nd Tampere Conference on Disaster Communications, 2001. 13. Nicholas Yeadon, Nigel Davies, Adrian Friday, Gordan Blair, “Supporting video in heterogeneous mobile environments“, Proceedings of the 1998 ACM symposium on Applied Computing, February 1998. 14. “Enabling high speed data communications over TETRA”, NOKIA white paper, 0205 PMR, 2005. 15. Pekka Blomberg, “The Pan-European trunking standard TETRA”, http://www.tetramou.com, 1997. 16. Wenlong Xu Haige Xiang Hongjie Yang, “TETRA protocol interfaces features and potential applications in railway”, TENCON ‘02, IEEE Region 10 Conference on Computers, Communications, Control and Power Engineering, pp. 1086 – 1088, vol.2, Oct. 2002. 17. Shiga BakariC, Mario Borzik, Darko BratkoviC, Vinko Grga, “TETRA (Terrestrial Trunked Radio) - Technical Features and Application of Professional Communication Technologies in MobiIe Digital Radio Networks for Special Purpose Services”, 47th International Symposium ELMAR, Zadar, Croatia, June 2005.

References

189

18. Martin Steppler, “TETRIS—A simulation tool for TETRA systems”, Mobile Kommunikation, Vol. 135, pp. 403– 410, September 1995. 19. L. Thornton, M. Chakraborty, J. Soraghan, “VIDEO OVER TETRA”, IEE Colloquium on Tetra Market and Technology Developments, London, February 2000. 20. Christian Hoymann, Dirk Kuypers, Peter Sievering, Peter Stuckmann, Bernhard Walke, Bangnan Xu, “Performance Analysis of TETRA and TAPS and Implications for Future Broadband Public Safety Communication Systems”, Workshop on Broadband Wireless Ad-Hoc Networks and Services, ETSI, Sophia Antipolis, France, September 2002. 21. Wang Zhou, Meissner Andreas, Grimmer Jan, Richter Stefanie, “A Mobile Multimedia Emergency Response Information System for Public Safety Organizations”, 2nd International Conference on Advances in Mobile Multimedia MOMM2004, pp 319-320, Bali, Indonesia, September 2004. 22. A Grilo, M Nunes, A Casaca, RA Redol, F Presutto, I Rebelo, “Communication Network Architecture for Mobiles Surveillance in an Airport Environment”, JISSA JOURNEES INTERNATIONALES SUR LES SENSEURS ET SYSTEMES DE SURVEILLANCE AEROPORTUAIRE, Paris, June 2005. 23. Marius Minea, Florin Codrut Nemtanu, Valentin Stan, “Establishing Communications Needs for the Urban Traffic and Public Transport Integrated System in Bucharest”, IEEE TELSIKS, Serbia and Montenegro, Nis, September 2005. 24. Apostolis Salkintzis, “Evolving Public Safety Communication Systems by Integrating WLAN and TETRA Networks”, IEEE Communications Magazine, January 2006. 25. Martin Steppler, “Maximum Number of Users Which Can Be Served by TETRA Systems”, Tagungsband der European Wireless ‘99, vol. 157, pp. 315-320, Munich, October 1999. 26. Alnuweiri, H., Kossentini, F. and Erol, B., “Efficient Coding and Mapping Algorithms for Software-only Real-Time Video Coding at Low Bit Rates”, IEEE Transactions on Circuits and Systems for Video Technology, no. 6, pp. 843-856 September 2000. 27. Côte, G., Erol, B., Gallant, M. and Kossentini, F., “H.263+: Video Coding at Low Bit Rates”, IEEE Transactions on Circuits and Systems for Video Technology, pp. 849-866, November 1998. 28. Cherriman P., “H.263 Video Coding”, online.:http://www.mobile.ecs.soton.ac.uk/ peter/h263/h263.html, 1999. 29. ITU-T Recommendation H.263 (1998). Video coding for low bit rate communication. 30. Côte, G., Kossentini, F. and Shirani, S. (2000). IEEE Journal on Selected Areas in Communications, vol. 18, no. 6, pp. 952-965, June 2000. Optimal mode selection and synchronization for robust video communications over error prone networks. 31. Martins, M. and Teixeira, L. (1996). European Conference on Multimedia Applications, Services and Techniques, Proceedings Part II, pp. 615-634,

190

32. 33. 34. 35. 36. 37. 38. 39. 40.

41. 42.

43.

44.

45.

5 Video Transmission over TETRA Louvain-la-Neuve, Belgium, May 1996. Video Compression: The MPEG Standards. ISO/IEC 13818-2 (2000). Information technology – Generic coding of moving pictures and associated audio information: Video Lee, H. (1997). Standard Coding for MPEG-1, MPEG-2 and Advanced Coding for MPEG-4 Online. Available: http://citeseer.nj.nec.com/lee97standard.html Chiariglione, L. (1998). MPEG-4, why use it? Online. Available: http://www.cselt.it/leonardo/paper/lcpaper.htm Zhang, Y. (1997). ISCAS Tutorial on MPEG-4. Chapter 3.4 Online.Available: http://bs.hhi.de/mpeg-video/contrib/iscas97/ ISO/IEC 14496-2 (2000). Information technology – Coding of audio-visual objects – Part 2: Visual Li, W. (1997). ISCAS Tutorial on MPEG-4. Chapter 3.2Online. Available: http://bs.hhi.de/mpeg-video/contrib/iscas97/ Brailean, J. (1997). ISCAS Tutorial on MPEG-4. Chapter 3.3 Online. Available: http://bs.hhi.de/mpeg-video/contrib/iscas97/ Samarakoon MI,Honary B,Rayne M.Encrypted video over TETRA.Tetra Market and Technology Developments (Ref.No.2000/007),IEE Seminar on10 Feb.2000 Tsagkaropoulos M., Politis I., Dagiuklas T., Kotsopoulos S., “Provisioning of Multimedia Applications across Heterogeneous All-IP Networks: Requirements, Functions and Research Issues”, Encyclopedia of Mobile Computing and Commerce vol. 2, 2006. Politis I., Dagiuklas T., Tsagkaropoulos M., Kotsopoulos S., “Interworking Architectures of 3G and WLAN towards All-IP Architectures: Comparisons”, Encyclopedia of Mobile Computing and Commerce vol.1, 2005. Politis I., Tsagkaropoulos M., Dagiuklas T., Kotsopoulos S., and Stavroulakis P., “On the Improvement of the Performance of Video Transmission over Integrated hybrid 3G-WLAN Networks”, Global Mobile Congress 2006 Special Workshop on 2008 China Olympic Games, October 2006, Beijing, China. I. Politis, M. Tsagkaropoulos, T. Dagiuklas and S. Kotsopoulos, “Study of the QoS of Video Traffic over Integrated 3G-WLAN systems”, 2nd International Mobile Multimedia Communications Conference, Alghero - Sardinia, Italy, September 18-20, 2006. Tsagkaropoulos M., Politis I., Dagiuklas T., Stylianakis V., and Kotsopoulos S., “On the Establishment of Dynamic Security and Trust Relations among Next Generation Heterogeneous Networks”. 45th FITCE Conference, Athens, August 2006. Dunlop J., Girma D., Irvine J., “Digital Mobile Communications and the TETRA System”, John Wiley & Sons Ltd., May 2000.

6 TETRA as a Gateway to Other Wireless Systems

Apostolis K. Salkintzis, Dimitrios I. Axiotis

6.1 Introduction In this chapter, we present an overview of packet data transmission over Terrestrial Trunked Radio (TETRA) release 1 networks as well as a solution for integrating TETRA with WLANs as a way to improve the packet data transmission capabilities. We first give a brief overview of the TETRA air interface and the available logical and physical channels. We then present various aspects of packet data transmission over TETRA, where we conclude that TETRA release 1 cannot provide the means to support demanding IP-based applications, mainly due to bandwidth and QoS constrains. Motivated from this conclusion, we then present a solution for integrating TETRA with Wireless Local Area Networks (WLANs) and thus realizing hybrid broadband networks suitable to support the next generation of public safety communication systems. The specified solution allows TETRA terminals to interface to the TETRA Switching and Management Infrastructure (SwMI) over a broadband WLAN radio access network, instead of the conventional narrowband TETRA radio network. These terminals are fully interoperable with conventional TETRA terminals and can employ all TETRA services, including group calls, short data messaging, packet data, etc. In addition, however, such terminals can support a range of brand new capabilities enabled by the WLAN, such as broadband data services, true concurrent voice and data services, simultaneous reception of many group calls, reduced call setup and voice transmission delays, improved voice quality, etc. It will become evident that the specified TETRA/WLAN integration solution is solely based on IP multicast and Voice-over-IP (VoIP) technologies and can thus fit ideally to the

192


all-IP architecture being introduced by the MESA project for the next generation of public safety and disaster relief communication systems.

6.2 TETRA Air Interface: Logical and Physical Channels The access scheme in TETRA is Time Division Multiple Access (TDMA) with 4 physical channels per carrier, and carrier separation is 25 kHz. The basic radio resource is a time slot lasting 14.167 ms (85/6 ms) transmitting information at a modulation rate of 36 kbit/s. The time slot duration, including guard and ramping times, is 510 bit (255 symbol) durations 1. The physical content of a time slot is carried by a burst. Additionally, a multiframe structure of 18 frames is defined that allows the introduction of associated control channels together with their corresponding traffic channels. A hyperframe is also defined, to facilitate the monitoring of adjacent cells by the mobile and accommodate a cryptographic scheme. The hyperframe level defines the top-level frame hierarchy. One hyperframe is subdivided into 60 multiframes, and lasts 61.2 s. One multiframe is subdivided in 18 frames, and has duration of 1.02 s. The eighteenth frame in a multiframe is a control frame. One frame is subdivided into 4 time slots, and has a duration of 170/3 ≈56.67 ms. The modulation scheme is π/4shifted Differential Quaternary Phase Shift Keying (π/4-DQPSK). In order to adequately reject adjacent channel power, limit intersymbol interference and facilitate receiver synchronization a root raised cosine filter is employed with a roll-off factor of 0.35 1. The TETRA air interface layer two is divided in two sublayers, namely Medium Access Control layer (MAC) and Logical Link Control layer (LLC). The main functions of the MAC relate to the preparation of data for transfer through radio connections and the associated control procedures. These functions can be further broken down into transforming of different types of information into TETRA TDMA structures, TDMA synchronization, error protection, encryption, signal strength, error rate measurement and random access procedures. The LLC handles data transmission, retransmission, segmentation re-assembly and logical link management and control. On top of LLC, in layer three, is the Mobile Link Entity (MLE) that is responsible for the control of lower layer radio functions in order to meet the requirements form the higher layers, and to isolate the higher layers from the activities of the radio link 1. The MLE performs the

6.2 TETRA Air Interface: Logical and Physical Channels

193

management of the radio connection, the management of mobility, identity management and Quality of Service mechanisms. 6.2.1 Logical Channels The radio subsystem provides a certain number of logical channels that represent the interface between the protocol and the radio. A logical channel is defined as a logical communication pathway between two or more parties. The logical channels may be separated into two categories: traffic channels carrying speech or data information in circuit switched mode and control channels carrying signaling messages and packet data.

Traffic Channels The traffic channels carry user information. Different traffic channels are defined for speech or data applications and for different data message speeds 1: • Speech Traffic Channel (TCH/S) • Circuit mode traffic channels: 7.2 kbit/s net rate (TCH/7.2), 4.8 kbit/s net rate (TCH/4.8), 2.4 kbit/s net rate (TCH/2.4) Higher net rates up to 28.8 kbit/s, 19.2 kbit/s or 9.6 kbit/s may be obtained by allocating up to 4 channels to the same communication.

Control CHannels (CCH) The CCH carry signaling messages and packet data. Five categories of control channels are defined 1: • Broadcast Control CHannel (BCCH) is a uni-directional channel for common use by all mobile Stations (MSs). It broadcasts general information to all MSs. Two subcategories are defined: o Broadcast Network Channel (BNCH) is downlink only and broadcasts network information to MSs. o Broadcast Synchronization Channel (BSCH) is downlink only and broadcasts information used for time and scrambling synchronization of the MSs. • Linearization CHannel (LCH) is used by the base and MSs to linearize their transmitter and two categories are defined: o Common Linearization Channel (CLCH) in the uplink, shared by all the MSs o BS Linearization CHannel (BLCH) in the downlink, used by the BS.

194


•

• •

Signaling CHannel (SCH) is shared by all MSs, but may carry messages specific to one MS or one group of MSs. System operation requires the establishment of at least one SCH per BS. SCH may be divided into 3 categories, depending on the size of the message: o Full size Signaling Channel (SCH/F) is a bidirectional channel used for full size messages o Half size Downlink Signaling Channel (SCH/HD) in the downlink only, is used for half size messages o Half size Uplink Signaling Channel (SCH/HU) in the uplink only, is used for half size messages Access Assignment CHannel (AACH) is present on all transmitted downlink slots. It is used to indicate on each physical channel the assignment of the uplink and downlink slots. STealing CHannel (STCH) is associated to a TCH that temporarily “steals” a part of the associated TCH capacity to transmit control messages. It is used when fast signaling is required.

6.2.2 Physical channels A physical channel is defined by a pair of radio carrier frequencies (downlink and uplink) and a Timeslot Number (TN). There shall be four physical channels per pair of radio frequencies. Three types of physical channel are defined 1: •

• •

Control Physical channel (CP) is a physical channel carrying exclusively CCH. Two types of CP channels are defined: o Main Control CHannel (MCCH). In each cell one RF carrier shall be defined as the main carrier. Whenever a MCCH is used, it is located on the timeslot 1 of the main carrier. o Secondary Control CHannel (SCCH). The SCCH may be used to extend the signaling capacity of the MCCH and may only be assigned when the MCCH is used. Traffic Physical channel (TP) is a physical channel carrying TCH. Unallocated Physical channel (UP) is a physical channel not allocated to one or more MS.

Further information on the mapping of the logical channels into physical channels and different types of bursts as well as the error control schemes for each logical channel, can be found in 1.

6.3 TETRA Packet Data Transmission

195

6.3 TETRA Packet Data Transmission Packet data transmission in TETRA takes place over an assigned secondary control channel (assigned SCCH), termed as Packet Data CHannel (PDCH). The assigned SCCH may be used by a certain group of MSs for a particular signaling message or data exchange (packet data), or it may be shared between several MSs each with intermittent bursts of signaling to send. The BS may use an assigned SCCH as a general packet data channel supporting advanced links for several MSs, where each MS may be intermittently offering data packets. An advanced link is set up before data transfer may begin on the PDCH. When a MS has data to transfer, it implicitly requests permission to switch to the PDCH. If accepted, the SwMI responds with a channel allocation, directing the MS to a PDCH. The Packet Data (PD) bearer service is the implementation of IP version 4 on top of the TETRA SubNetwork Dependent Convergence Protocol (SNDCP), see Figure 1. The PD bearer service is terminated at the IP protocol. The Packet Data service follows the principles used in the IP world. Delivery is on a best effort basis, and if it cannot be achieved (MS not registered, MS out of range etc.) the datagram is discarded. The system supports the Internet Control Message Protocol (ICMP), which means that any non-deliveries will be reported to the source (ICMP messages). The PD service is designed to take care of all mobility aspects for the users of the service, thus a host connected to a MS does not need to know the location of the destination when sending a datagram. The applications however, need to take into account the narrow bandwidth channel and varying environmental conditions as the bandwidth and delay may vary with signal conditions and network loading. Each message is sent across the air interface using the TETRA SNDCP service. The underlying TETRA layer 2 (LLC & MAC) splits the message up into segments which are mapped into slots. Since each segment carries a fixed amount of bits, the used air interface resources increase in jumps on some message length boundaries. The “cost” of sending an IP datagram is therefore a basic signaling cost, plus a “variable cost” depending on the size of the IP datagram. The boundaries where additional slots are necessary are summarized in Table 1 2.

196


Figure 1. Usage of TETRA packet data for TE IP application Table 1. Slots needed in the uplink/downlink vs. Max IP datagram size AdditionalMax IP AdditionalMax IP Additional Max IP Additional Max IP Uplink datagram Uplink datagram Downlink datagram Downlink datagram Slots size Slots size Slots size Slots size (bytes) (bytes) (bytes) (bytes) -

50

11

364

-

49

11

354

1

79

12

392

1

76

12

382

2

107

13

421

2

104

13

409

3

136

14

449

3

132

14

437

4

164

15

478

4

160

15

465

5

193

16

506

5

187

16

493

6

221

17

535

6

215

17

520

7

250

18

563

7

243

18

548

8

278

19

592

8

271

19

576

9

307

20

620

9

298

20

604

10

335

21

649

10

326

21

631


197

For example in the downlink, any IP datagram from 21 (minimum IP datagram size) to 49 bytes will consume the same amount of resources. Note that the IP datagram size includes the IP and transport protocol header. For UDP, the IP and UDP overhead is 28 bytes. In order to send 130 bytes of user data, an IP datagram of length 158 bytes must be sent this requires 4 additional downlink slots. The PDCH may offer in good conditions an instantaneous throughput around 3 kbp/s 2. This bandwidth will possibly have to be shared with other users using the same PDCH. In addition, RF conditions and PDCH load will cause the delay to vary significantly. Typical best case “ping” round trip delays are in the order of 1s, but the use of longer datagrams, RF conditions and PDCH load can make this number increase substantially. TETRA packet data extends TETRA to act as an IP subnet. This enables application programmers to build their applications in a well-standardised environment. In 3, we performed a UDP performance measurement survey for a commercially deployed TETRA network which provides dense coverage in major Greek cities and autoroutes, through over 90 Base Stations (being one of the largest networks in Europe). The network is Voice+Data (V+D) ready and one PDCH was available in the cells where we performed the measurements. The network served 16.000 of users during the Athens 2004 Olympic Games. The frequency range allocated to TETRA network is 413.750 – 415.725 MHz (uplink) and 423.750 – 425.725 MHz (downlink). The mean throughput 3 is nearly an increasing function of the datagram size for the range between 15 and 180 bytes and varies between 0.4 and 1.8 kbit/s. It then remains roughly constant between 1.8-1.9 kbit/s for the range 180-285 bytes and finally decreases between 285-600 bytes, varying between 1.7 and 1.2 kbit/s. During the measurements, the instantaneous throughput reached 3kbit/s for the range 180-285 bytes, thus confirming the theoretical limit given in 2. It is obvious that for the smaller datagram sizes (below 80 bytes) the UDP payload is comparable to the IP overhead (20 bytes) thus resulting in low throughput. It seems that the optimal datagram size in terms of system throughput ranges between 125-285 bytes, since the mean throughput is over 1.5kbps in all occasions 3. Results we obtained include throughput and delay jitter (mean values and standard deviations), percentage of lost datagrams and percentage of datagrams received out-of-order versus the datagram sizes; further information can be found in 3.

198


6.3.1 Packet Data transmission and reception procedures In this section we shall describe the procedures within SNDCP for transmitting and receiving packet data. The basic setting for sending N-PDUs between an MS and the SwMI is the acknowledged service. A scenario illustrating acknowledged data sending from SwMI to MS is shown in Fig. 2. Each numbered step is explained in the following list.

Fig. 2. Packet data transfer from SwMI to MS

1. The SwMI SNDCP receives the first SN-DATA request primitive. 2. The SwMI SNDCP sends SN-DATA TRANSMIT REQUEST PDU. The request is sent by using the acknowledged or unacknowledged service (basic link) and the MAC-RESOURCE PDU contains a “Channel Allocation” information element sending the MS to a PDCH. On transmission of the SN-DATA TRANSMIT REQUEST PDU, the SwMI stops the STANDBY timer, starts the READY timer and enters state READY. On reception of the SN-DATA TRANSMIT REQUEST PDU, the MS stops the STANDBY timer, starts the READY timer,


199

starts the CONTEXT_READY timer for the PDP context referenced by the NSAPI parameter in the SN-DATA request primitive, and enters state READY. 3. The SwMI SNDCP receives the second SN-DATA request primitive. 4. The first N-PDU is sent to the MS by issuing SN-DATA PDU. On reception of an indication from the MLE, in the form of an MLEREPORT indication primitive, that the SN-DATA PDU was successfully transmitted by the lower layers, the SwMI SNDCP entity shall restart the READY timer. At the MS the reception of the SN-DATA PDU shall result in the READY timer and, if supported, the CONTEXT_READY timer being restarted. 5. After receiving the first SN-DATA PDU the MS SNDCP passes the N-PDU to the higher layer by issuing a SN-DATA indication primitive 6. After receiving acknowledgement that the first SN-DATA PDU was successfully sent to the MS the SwMI sends SN-DELIVERY indication primitive to the higher layer. 7. The second N-PDU is sent to the MS by issuing SN-DATA PDU. The READY timer is restarted in the SwMI SNDCP and in the MS SNDCP. The relevant CONTEXT_READY timer is restarted in the MS SNDCP. 8. After receiving the second SN-DATA PDU the MS SNDCP passes the N-PDU to the higher layer by issuing a SN-DATA indication primitive. 9. After receiving acknowledgement that the second SN-DATA PDU was successfully sent to the MS the SwMI sends SN-DELIVERY indication primitive to the higher layer. 10. The SwMI SNDCP receives the third SN-DATA request primitive. 11. The third N-PDU is sent to the MS by issuing SN-DATA PDU. The READY timer is re-started in the SwMI SNDCP and in the MS SNDCP. The relevant CONTEXT_READY timer is restarted in the MS SNDCP. 12. After receiving the third SN-DATA PDU the MS SNDCP passes the N-PDU to the higher layer by issuing SN-DATA indication primitive. 13. After receiving acknowledgement that the third SN-DATA PDU was successfully sent to the MS the SwMI sends SN-DELIVERY indication primitive to the higher layer.

200


14. The READY timer expires in the SwMI SNDCP and it issues SNEND OF DATA PDU, including channel allocation, to the MS, starts the STANDBY timer and enters STANDBY state. The MS SNDCP receives SN-END OF DATA PDU, stops its READY timer, starts the STANDBY timer and enters STANDBY state on the indicated channel. A scenario illustrating acknowledged data sending from MS to SwMI is shown in Fig. 3. Each numbered step is explained in the following list.

Fig. 3. Packet data transfer from MS to SwMI

1. The MS SNDCP receives the first SN-DATA request primitive. 2. The MS SNDCP sends SN-DATA TRANSMIT REQUEST PDU. The request is sent by using acknowledged service (basic link). On sending the SN-DATA TRANSMIT REQUEST PDU, the MS SNDCP entity, starts the RESPONSE_WAIT timer and enters RESPONSEWAITING state. 3. The MS SNDCP receives the second SN-DATA request primitive. 4. The SwMI SNDCP sends SN-DATA TRANSMIT RESPONSE PDU. The response is sent by using the acknowledged or unacknowledged service (basic link) and the MAC-level MAC-RESOURCE PDU contains channel allocation information element commanding the MS to PDCH. On transmission of the SN-DATA TRANSMIT RESPONSE PDU, the SwMI stops the STANDBY timer, starts the READY timer and enters state READY. On reception of the SN-DATA TRANSMIT RESPONSE PDU, the MS stops the STANDBY and


201

RESPONSE_WAIT timers, starts the READY timer and the relevant CONTEXT_READY timer, and enters state READY. 5. The MS SNDCP receives the third SN-DATA request primitive. 6. The first N-PDU is sent to the SwMI by issuing SN-DATA PDU. On reception of an indication from the MLE, in the form of an MLEREPORT indication primitive, that the SN_DATA PDU was successfully transmitted by the lower layers, the MS SNDCP entity shall restart the READY timer and shall, if supported, restart the relevant CONTEXT_READY timer. At the SwMI the reception of the SNDATA PDU shall result in the READY timer being restarted. 7. After receiving acknowledgement that the first SN-DATA PDU was successfully sent to the SwMI, the MS sends SN-DELIVERY indication primitive to the higher layer. 8. After receiving the first SN-DATA PDU, the SwMI SNDCP passes the N-PDU to the higher layer by issuing SN-DATA indication primitive. 9. The second N-PDU is sent to the SwMI by issuing SN-DATA PDU. The READY timer is restarted in the MS SNDCP and in the SwMI SNDCP. The relevant CONTEXT_READY timer is restarted in the MS SNDCP. If scheduled access was agreed with the SwMI for this PDP context during PDP context activation, the MS SNDCP indicates that the SN-PDU is scheduled and includes the value of the maximum schedule period for this PDP context in the MLE-UNITDATA request primitives carrying this and all subsequent SN-DATA PDUs for this PDP context. 10. After receiving acknowledgement that the second SN-DATA PDU was successfully sent to the SwMI, the MS sends SN-DELIVERY indication primitive to the higher layer. 11. The third N-PDU is sent to the SwMI by issuing SN-DATA PDU. The READY timer is re-started in the MS SNDCP and in the SwMI SNDCP. The relevant CONTEXT_READY timer is restarted in the MS SNDCP. 12. After receiving the second SN-DATA PDU, the SwMI SNDCP passes the N-PDU to the higher layer by issuing SN-DATA indication primitive.

202


13. After receiving acknowledgement that the third SN-DATA PDU was successfully sent to the SwMI, the MS sends SN-DELIVERY indication primitive to the higher layer. 14. After receiving the third SN-DATA PDU the SwMI SNDCP passes the N-PDU to the higher layer by issuing SN-DATA indication primitive. 15. The READY timer expires in the SwMI SNDCP and it issues SNEND OF DATA PDU, including channel allocation, to the MS, starts the STANDBY timer and enters STANDBY state. The MS SNDCP receives SN-END OF DATA PDU, stops its READY timer, starts the STANDBY timer and enters STANDBY state in the allocated channel. The READY timer function maintains the READY timer in the MS and SwMI. The READY timer may be defined for each MS separately. The READY timer controls, with the help of other potential timers, the time an MS and SwMI remains in READY state after either a SN-DATA, SNUNITDATA, SN-DATA TRANSMIT REQUEST PDU (SwMI to MS) or SN-DATA TRANSMIT RESPONSE PDU (SwMI to MS) has been transmitted between the MS and SwMI. When the READY timer expires in the MS, the MS shall send SN-END OF DATA PDU to the SwMI and restart READY timer. The MS shall stop the READY timer, start the STANDBY timer and enter to the STANDBY state on receiving SN-END OF DATA PDU. When the READY timer expires in the SwMI, the SwMI shall send a SN-END OF DATA PDU to the MS and enter to the STANDBY state. The SwMI may also send a SN-END OF DATA PDU to the MS for its own reasons. 6.3.2 TETRA IP user authentication Fig. 4 illustrates the reference model of IP user authentication when using PPP and RADIUS protocols (IETF RFC 1661 4, IETF RFC 2865 6). In the model an AAA server and the RADIUS protocol are used to verify the user access. Other alternatives for the model are also possible. For instance inside the SwMI there could be a user authentication entity to provide the same functionality as the external AAA server.


203

Fig. 4. IP user authentication model

Following assumptions are made: • • • • • •

PPP is the link layer protocol used between a TETRA Terminal Equipment (TE) and a TETRA Mobile Termination (MT); there is a requirement to authenticate the TE using PAP or CHAP, refer to IETF RFC 1994 7; the TE is the peer that shall be authenticated, and the MT is the authenticator, using the terminology defined in ISO/IEC 8348 8; there is a requirement to support the PPP authentication with a centralised AAA server which is accessed by RADIUS protocol as defined in RFC 2865 6; the PAP or CHAP authentication information collected in the MT is forwarded over the TETRA Air Interface to the TETRA SwMI; inside the TETRA SwMI is a RADIUS client entity which forwards the authentication information to the external AAA server.

Fig. 5 illustrates the phases of a packet data context setup upon a successful authentication with CHAP. Corresponding signaling using PAP authentication would be slightly more straightforward.

204


Fig. 5. A successful authentication with CHAP

The following steps clarify Fig. 5: 1. PPP/LCP negotiates the Maximum-Receive-Unit and authentication options; 2. When using CHAP (or PAP), the TE authenticates itself to the MT, which stores the information (username, challenge and response, or password when using PAP) and sends an accept to TE; 3. TE requests IP configuration from the MT using PPP/IPCP, either defining a static address, or requesting an address from the network; 4. MT sends an ACTIVATE PDP CONTEXT DEMAND PDU to SwMI, containing in the Protocol configuration option information element the authentication and configuration information it has collected; 5. RADIUS client in SwMI sends an Access-Request to AAA Server using RADIUS protocol; 6. AAA Server sends an Access-Accept to RADIUS Client; 7. SwMI sends an ACTIVATE PDP CONTEXT ACCEPT PDU to MT; 8. MT sends an IPCP Configure Ack to TE and the link is open (or dropped if negotiation failed).

6.4 SNDCP states and state transitions

205

6.4 SNDCP states and state transitions The SNDCP activities related to a TETRA MS are characterised by one of eight different SNDCP states: CLOSED, IDLE, IDLE-Temporary Break, STANDBY, STANDBY-Temporary Break, RESPONSEWAITING, READY and READY-Temporary Break. The SNDCP activities related to a TETRA SwMI are characterised by one of three different SNDCP states: IDLE, STANDBY and READY. Each state describes a certain level of functionality and information allocated to the involved entities. The SNDCP state relates only to SNDCP activities of a subscriber represented by the Individual TETRA Subscriber Identity (ITSI). It is independent of number of Packet Data Protocol (PDP) contexts for that subscriber. It is optional for a MS to support multiple PDP contexts. It is also optional for a SwMI to support multiple PDP contexts for a single ITSI.

CLOSED

CLOSED state is valid for a MS only. In CLOSED state access to the communication resources is unavailable (e.g. due to MS not being registered) and SNDCP is not permitted to communicate with its peer entity. In CLOSED state, the MS must not have any PDP contexts active. When entering state CLOSED, the READY, CONTEXT_READY, RESPONSE_WAIT and STANDBY timers are stopped. On reception of an indication that access to the communication resources has become available (on reception of MLE-OPEN indication primitive from MLE), the MS SNDCP entity shall enter IDLE state.

IDLE

In IDLE state the MS and SwMI shall not have PDP contexts. When entering to state IDLE, the STANDBY, RESPONSE_WAIT, READY, and CONTEXT_READY timers are stopped. Data transfer to and from the mobile subscriber is not possible. The MS is seen as not reachable in this case for TETRA Packet data. In order to establish SNDCP contexts in the MS and the SwMI, the MS shall perform the PDP context activation procedure. After successful PDP context activation the MS shall start STANDBY timer and enter to state STANDBY. On reception of an MLEBREAK indication primitive from the MLE, the MS SNDCP entity shall enter IDLE-Temporary Break state.

206


IDLE-temporary break

IDLE-Temporary Break state is only valid for the MS. This state is only entered while access to the communication resources has become temporarily unavailable (e.g. due to cell reselection). A temporary break in access to the communication resources is signalled to SNDCP by reception of the MLE-BREAK indication primitive from the MLE. This state shall be entered from IDLE state on reception of an MLE-BREAK indication primitive. This state shall also be entered from STANDBY-Temporary Break state when the STANDBY timer expires. On entering this state from STANDBY-Temporary Break, all PDP contexts shall be locally deactivated. In IDLE-Temporary Break state the MS and SwMI shall not have PDP contexts. Communication between the MS and SwMI SNDCP entities is not possible in this state. On reception of an MLE-RESUME indication primitive from the MLE, the MS SNDCP entity shall enter IDLE state. STANDBY

In STANDBY state, the subscriber has at least one PDP context activated. The MS may receive and respond to SN-PAGE REQUEST PDUs while in this state. The MS may initiate activation of a new PDP context while in STANDBY state. The MS and SwMI may initiate modification of PDP contexts while in STANDBY state. The MS may initiate deactivation of PDP contexts while in STANDBY state. After deactivation of the last PDP context assigned by the MS to a particular layer 2 logical link, the MS SNDCP entity shall issue an MLEDISCONNECT request primitive to the MLE in order to disconnect that logical link. After deactivation of the last PDP context the STANDBY timer is stopped and SNDCP state is changed to IDLE. If the MS SNDCP has previously informed the MLE that the SNDCP status is “standing-by” or “ready” and the MS no longer expects to transmit or receive packet data in the forseeable future, the MS SNDCP should inform the MLE that its SNDCP status is now “idle” (using the MLE-CONFIGURE request primitive). The MS MLE may use this information to stop assessing channel classes and monitoring sectored channels. The information should be sent to the MS MLE when SNDCP returns to the IDLE state, but may be sent sooner.


207

On reception of a SN-DATA request or SN-UNITDATA request primitive from the service user when the SERVICE_CHANGE timer is inactive, the MS SNDCP entity shall transmit a SN-DATA TRANSMIT REQUEST PDU. On transmission of a SN-DATA TRANSMIT REQUEST PDU, the MS SNDCP entity shall send an MLE-ACTIVITY request primitive with sleep mode set to “stay alive”, start the RESPONSE_WAIT timer and enter RESPONSE-WAITING state. The STANDBY timer is not stopped on entering RESPONSE-WAITING state. On reception (MS only) or transmission (SwMI only) of a SNACTIVATE PDP CONTEXT ACCEPT PDU or a SN-MODIFY PDP CONTEXT RESPONSE PDU the SNDCP entity shall restart the STANDBY timer. The STANDBY timer is stopped and the SNDCP state is changed to READY on transmission of (SwMI only) or reception of (MS only) SN-DATA TRANSMIT REQUEST PDU. Upon moving to the READY state the MS SNDCP shall issue an MLE-ACTIVITY request primitive with sleep mode set to “stay alive”. On transmission of a SN-DATA TRANSMIT RESPONSE PDU with Accept/Reject = 1 (i.e. Request accepted), the SwMI SNDCP entity shall stop the STANDBY timer, start the READY timer and enter READY state. On transmission of a SN-DATA TRANSMIT RESPONSE PDU with Accept/Reject = 0 (i.e. Request rejected), the SwMI SNDCP entity shall remain in STANDBY state. On reception of an SN-UNITDATA PDU, the MS SNDCP shall enter the READY state and, if the PDU was individually addressed, shall start the READY timer and, if supported, a CONTEXT_READY timer. Where there is a temporary break in access to the radio communication resources as indicated by the reception of an MLE-BREAK indication primitive from the MLE, the MS SNDCP entity shall issue MLERELEASE request primitives asking MLE to locally disconnect the advanced links and shall enter STANDBY-Temporary Break state.

STANDBY-temporary break

STANDBY-Temporary Break state is only valid for the MS. This state is only entered while access to the communication resources has become temporarily unavailable (e.g. due to cell reselection). A temporary break in access to the communication resources is signalled to SNDCP by reception of the MLE-BREAK indication primitive from the MLE. This state shall be entered from STANDBY state and RESPONSE-WAITING state on

208


reception of an MLE-BREAK indication primitive. This state shall also be entered from READY-Temporary Break state when the READY timer expires. Communication between the MS and SwMI SNDCP entities is not possible in this state. On reception of an MLE-RESUME indication primitive from the MLE, the MS SNDCP entity shall enter STANDBY state. After successful PDP context activation the MS shall start the STANDBY timer and enter to state STANDBY. If the STANDBY timer expires, the PDP contexts are deleted locally and the SNDCP state is changed to IDLETemporary Break. The SNDCP entity shall issue a SN-NSAPI DEALLOC indication primitive to the service user having “Deactivation type” parameter set to value “Deactivate all NSAPIs”.

RESPONSE-WAITING

RESPONSE-WAITING state is only valid for the MS. In RESPONSEWAITING state, the MS has at least one PDP context activated. The MS SNDCP entity shall enter RESPONSE-WAITING state from STANDBY state on transmission of a SN-DATA TRANSMIT REQUEST PDU. On entering RESPONSE-WAITING state the STANDBY timer remains active and the RESPONSE_WAIT timer is started. The MS SNDCP entity shall also enter RESPONSE-WAITING state from READY-Temporary Break state on reception of an MLE-RESUME indication primitive from the MLE if pending SN-DATA request or SN-UNITDATA request primitives from the service user cause the MS to transmit an SN-RECONNECT PDU containing “Data to Send” = 1. In this case the MS stops the READY timer and starts the RESPONSE_WAIT and STANDBY timers. The MS shall not initiate the activation or modification of PDP contexts while in RESPONSE-WAITING state. The MS shall not initiate the deactivation of PDP contexts while in RESPONSE-WAITING state. The MS may respond to a SN-PAGE REQUEST while in RESPONSE-WAITING state. On reception of a SN-DATA request or SN-UNITDATA request primitive from the service user, the MS SNDCP entity shall store the request. On reception of a SN-DATA TRANSMIT REQUEST PDU, the MS SNDCP shall stop the STANDBY and RESPONSE_WAIT timers, start the READY timer and, if supported, the CONTEXT-READY timer for the indicated PDP context, and enter READY state. On reception of a SNDATA TRANSMIT RESPONSE PDU with Accept/Reject = 1, the MS SNDCP shall stop the STANDBY and RESPONSE_WAIT timers, start the READY timer and enter READY state. On reception of a SN-DATA TRANSMIT RESPONSE PDU with Accept/Reject = 0, the MS SNDCP


209

shall send an MLE-ACTIVITY request primitive with sleep mode set to “sleep permitted”; stop the RESPONSE_WAIT timer; if the MS SNDCP was awaiting a response to an SN-RECONNECT PDU: issue MLERELEASE request primitives instructing MLE to locally disconnect the advanced links; enter STANDBY state. Where the RESPONSE_WAIT timer expires, the MS SNDCP shallsend an MLE-ACTIVITY request primitive with sleep mode set to “sleep permitted”;if the MS SNDCP was awaiting a response to an SNRECONNECT PDU: issue MLE-RELEASE request primitives instructing MLE to locally disconnect the advanced links; enter STANDBY state. Where the STANDBY timer expires, the RESPONSE_WAIT timer is stopped, all PDP contexts are deleted locally and the SNDCP state is changed to IDLE. The SNDCP entity shall send an MLE-ACTIVITY request primitive with sleep mode set to “sleep permitted”, issue a SNNSAPI DEALLOC indication primitive to the service user having “Deactivation type” parameter set to value “Deactivate all NSAPIs”.

READY

In READY state, the subscriber has at least one PDP context activated. The MS may receive and transmit N-PDUs while in this state. The MS SNDCP shall enter READY state on reception of a SN-DATA TRANSMIT REQUEST PDU or of a SN-DATA TRANSMIT RESPONSE PDU (with Accept/Reject = 1) and the MS SNDCP shall send an MLE-ACTIVITY request primitive with sleep mode set to “stay alive”. On entering the READY state, the MS SNDCP shall stop the RESPONSE_WAIT and STANDBY timers. If the SN-DATA TRANSMIT REQUEST PDU was individually addressed (but not if it was group addressed), the MS SNDCP shall start the READY timer and, if supported, a CONTEXT_READY timer for the relevant PDP context (i.e. the PDP context whose NSAPI was given in the SN-DATA TRANSMIT REQUEST PDU or SN-DATA TRANSMIT RESPONSE PDU). The SwMI SNDCP shall enter READY state on transmission of a SN-DATA TRANSMIT RESPONSE PDU (with Accept/Reject = 1) or of a SNDATA TRANSMIT REQUEST PDU. On entering READY state, the RESPONSE_WAIT and STANDBY timers are stopped and the READY timer is started. In the case where the MS enters READY state after reception of a group addressed SN-DATA TRANSMIT REQUEST PDU, the MS shall not start the READY timer or a CONTEXT_READY timer.

210


After transmitting an SN-DATA TRANSMIT REQUEST PDU in the READY state, the MS SNDCP shall send an MLE-ACTIVITY request primitive with sleep mode set to “stay alive”, shall start (or restart) the RESPONSE_WAITING timer and remain in the READY state. It shall not transmit any SN-DATA or SN-UNITDATA PDUs referencing that NSAPI until it receives an SN-DATA TRANSMIT RESPONSE PDU referencing that NSAPI with Accept/Reject = 1. On receiving an SN-DATA TRANSMIT RESPONSE PDU with Accept/Reject = 1 referencing a waiting NSAPI, the MS SNDCP shall stop the RESPONSE-WAITING timer, shall set up a new advanced link if required, and shall commence transmitting SN-DATA PDUs or SN-UNITDATA PDUs for that NSAPI. On reception of an SN-DATA TRANSMIT RESPONSE PDU with Accept/Reject = 0, or if the RESPONSE-WAITING timer expires, the MS SNDCP shall stop the RESPONSE_WAIT timer and remain in the READY state. The MS SNDCP entity shall ensure that all stored SNDATA request and SN-UNITDATA request primitives relating to the relevant PDP context are deleted. For each SN-DATA request and SNUNITDATA request primitive deleted, a corresponding notification of failure shall be sent to the service user in the form of a SN-DELIVERY indication primitive. The MS shall not deactivate PDP contexts while in READY state. To initiate the deactivation of one or more PDP contexts, a MS must return to STANDBY state. The MS may initiate activation of a new PDP context while in READY state. The MS may also initiate modification of PDP contexts while in this state. Should either happen, the MS shall remain in the state READY. Except during cell change or following loss of radio resources, the MS SNDCP remains in the READY state while the READY timer is active even when there is no data being communicated. Where there is a temporary break in access to the radio communication resources as indicated by the reception of an MLE-BREAK indication primitive from the MLE, the MS SNDCP entity shall enter READY-temporary break state.

READY-temporary break

READY-Temporary Break state is only valid for the MS. This state shall be entered only from the READY state and only when access to the communication resources has become temporarily unavailable (e.g. due to cell reselection or loss of radio resource). A temporary break in access to the communication resources is signalled to SNDCP by reception of the

6.5 UDP versus TCP on top of TETRA IP layer

211

MLE-BREAK indication primitive or MLE-CONFIGURE indication primitive indicating “loss of radio resources” from the MLE. Communication between the MS and SwMI SNDCP entities may not be possible in this state. While in the READY-temporary break state, the MS SNDCP shall not attempt to transmit SN-DATA or SN-UNITDATA PDUs. When the MS SNDCP is in the READY-temporary break state, the MS should suspend transmission of N-PDUs, including any partially sent or unacknowledged N-PDUs in the LLC buffers, until the MS SNDCP returns to the READY state. On reception of an MLE-RESUME indication primitive from the MLE, the MS SNDCP entity shall first check if it has a pending SN-DATA request or SN-UNITDATA request primitive. If there is data awaiting transmission, then the MS SNDCP entity shall send to the SwMI a SN-RECONNECT PDU with the field “Data to Send” set to 1. It shall then stop the READY timer and any CONTEXT_READY timers, shall start the STANDBY and RESPONSE_WAIT timers and enter RESPONSE-WAITING state. If there is no data awaiting transmission, then MS SNDCP entity shall send to the SwMI a SN-RECONNECT PDU with the field “Data to Send” set to 0. It shall then stop the READY timer, start the STANDBY timer and enter STANDBY state. After successful PDP context activation the MS shall start the STANDBY timer and enter STANDBY state. On reception of a SNDATA TRANSMIT RESPONSE PDU with Accept/Reject = 1, the MS SNDCP shall stop the RESPONSE_WAIT timer, start the READY timer and, if supported, the CONTEXT_READY timer for the indicated PDP context and shall enter the READY state. On reception of an SN-DATA TRANSMIT RESPONSE PDU with Accept/Reject = 0, the MS SNDCP shall send an MLE-ACTIVITY request primitive with sleep mode set to “sleep permitted”, stop the RESPONSE_WAIT timer, the READY timer and any CONTEXT_READY timers, start the STANDBY timer and shall enter the STANDBY state.

6.5 UDP versus TCP on top of TETRA IP layer As will be analyzed in the following, the recommended transport protocol on top of IP is UDP. It is a minimal message-oriented transport layer protocol, documented in 4. Using UDP, applications running in Terminal

212


Equipment (TE1) or MS can send short messages known as datagrams to one another. UDP does not provide the reliability and ordering guarantees that Transmission Control Protocol (TCP) does, hence datagrams may arrive out-of-order or be lost without notice. However, as a result, UDP is faster and more efficient for many lightweight or time-sensitive purposes. UDP adds only application multiplexing and data checksumming on top of an IP datagram. Lacking reliability, UDP applications must generally be willing to accept some loss or errors. Some applications may implement additional reliability mechanisms in the application layer if needed. If an application requires a high degree of reliability, a protocol such as the TCP may be used instead. The network does not store data in case an IP datagram is not delivered; it drops the datagram and sends back an ICMP message to the originator. The originator must then decide what action to take. The returned ICMP message will provide the reason why the message was not delivered. For some applications, loss of a packet may not be critical. In the simplest usage of PD, the user sends a packet and ignores all reasons that cause a single packet to be lost. However, if the system refuses to deliver the message because the destination is unknown in the system, the application should flag an error to the user so that corrective action can be taken. If an ICMP “Host unreachable” message is received for a packet, the network has done its best to deliver the packet. In this case, retransmitting the packet will most likely not result in delivery success. Hence, the application should implement a kind of back-off and suspend sending packets for a while. Unlike UDP, TCP needs to establish a connection before any data can be exchanged. This means that several messages must be exchanged between the initiator (the client) and the receiver (the server), before the actual data is sent. TCP in this case has a large overhead compared to using UDP. TCP also retransmits segments for which it has not received acknowledgements within the retransmission timeout period (RTO). The first TCP retransmission occurs after 1*RTO, the next after 2*RTO, then 4*RTO, etc. and the procedure continues until the interval between retransmissions reaches a predefined upper limit. After this, all retransmissions will occur with an interval equal to this limit. Retransmissions will continue until TCP eventually gives up - in a typical TCP implementation this will happen after 9 minutes. The actual RTO used by the transmitter is an estimate of the round trip time (RTT). Typically, the RTO is set to

A TE is a computer running the TCP/IP protocol suite. It uses PPP to connect to the MS and IP v4 to communicate with other computers running the TCP/IP. That is, a TE is a host connected to an MS. 1

6.6 TETRA Packet Data modems 213

twice the estimated RTT. The retransmission strategy in TCP is designed for use in large IP networks, usually characterized by local links with plenty of bandwidth or long haul links shared by many users. Both of these will tend to provide an RTT variance that is less than the one experienced in a radio environment. As an example, if TCP adapts to the radio environment during one access of the PDCH, then the next time the PDCH is accessed, the traffic conditions may have changed radically (idle to busy system, good signal reception to bad reception).

6.6 TETRA Packet Data modems Interfacing to a MS to access the packet data service is very similar to the way you interface to a wireline modem. When configuring a TE (e.g. a PC) to interface with a MS, the TE will generally be configured to believe that it is connected to a wireline modem. However, there are some essential differences between a TETRA and a wireline modem. One major difference is that a wireline modem provides a circuit mode connection to the Network Access Server (NAS), whereas the Packet Data service is entirely packet switched. Thus the NAS, the Packet Data Gateway (PDG), and the MS communicate in a truly packet switched manner. Hence, if no data is being sent, no bandwidth is consumed. Furthermore, there is no need to tear down the communication with the NAS (PDG): the MS/TE can remain “connected” all the time, providing very fast access to the packet data service. Another major difference is that, in a normal wireline modem environment, the PPP link from the host is terminated in the NAS, i.e. the modem simply carries PPP transparently. In the packet data service, the PPP link is terminated in the MS. Whenever an IP datagram is sent to/from a host, the IP datagram is re-packed in a TETRA specific protocol, optimized for the radio communication environment. This adaptation to the radio environment provides the following advantages: •

•

Radio link layer failure may cause lost IP datagrams but will not cause lost dial-up calls. In a circuit switched dial-up environment, (radio) link layer failure will typically cause the circuit (call) to be dropped. Sessions can stay alive even if the radio temporarily moves out of coverage (the PPP link is not broken). In the circuit switched case, re-establishing the circuit requires that the host re-dials the connection and re-establishes the PPP link. This will typically cause all sessions to be lost, requiring the user to re-login and re-authenticate.

214


•

IP datagrams lost on the radio link layer can be reported as lost (ICMP).

6.6.1 Types of Packet-data Mobile Stations In this section we describe the service interactions between circuit mode, messaging and packet data services as defined by the value of the Packet data MS Type information element. It is used by the MS to indicate to the SwMI its service interaction capabilities between packet data and individual circuit mode speech and data calls. The Packet data MS Type is sent to the SwMI when activating PDP context. If the Packet data MS Type is changed after the first PDP context activation, then the SwMI shall use the Packet data MS Type value, as received in the latest PDP activation. The SwMI may use the Packet data MS type value to assist in optimizing air interface signaling when offering packet data services to an MS. For example the SwMI may choose not to offer (because the MS will probably not accept) TETRA Packet data to type D and C MS which is engaged in a circuit mode speech/data or Circuit mode speech/data to type D and C MS which is engaged in TETRA Packet data. The different types for TETRA Packet data MSs and their capabilities to handle other services in addition to TETRA Packet data are summarized heafter.

Type A - Parallel

All services (SDS, Circuit mode speech/data) may be conducted parallel with TETRA Packet data. MS has the capability to receive/send both circuit mode speech/data and TETRA Packet data at the same time. This means that circuit mode speech/data may be set up, accepted and conducted while the MS is in READY state. SDS may be conducted while the MS is in READY state.

Type B - Alternating

Other services (SDS, Circuit mode speech/data) may be conducted alternating with TETRA Packet data, i.e. if MS is engaged in TETRA Packet data it may accept circuit mode speech/data and vice versa but only one service (TETRA Packet data or circuit mode speech/data) can be active at any time. SDS can be conducted parallel both with TETRA Packet data

6.6 TETRA Packet Data modems 215

and circuit mode speech/data. This means that circuit mode speech/data may be set up but not conducted while the MS is in READY state. Circuit mode speech/data may be only conducted in STANDBY state.

Type C - IP single mode

The MS is engaged to one service at a time; if MS is engaged in TETRA Packet data it does not accept circuit mode speech/data and vice versa: 1. While the MS is in the READY state i.e. engaged in TETRA Packet Data, no circuit mode speech/data calls may be set up, accepted or conducted. However, SDS may be conducted while the MS is in the READY state. 2. While the MS has an ongoing circuit mode speech/data call or is in the call set up phase, TETRA Packet data is not allowed (accepted or initiated). This means that circuit mode speech/data may be set up, accepted or conducted while the MS is in STANDBY state. SDS may also be conducted while the MS is in the STANDBY state.

Type D - Restricted IP single mode

The MS is engaged to one service at a time; if MS is engaged in TETRA Packet data it does not accept circuit mode speech/data and does not support SDS and vice versa: 1. While the MS is in the READY state i.e. engaged in TETRA Packet Data, no circuit mode speech/data may calls be set up, accepted or conducted. SDS should not be conducted while the MS is in READY state. 2. While the MS has an ongoing circuit mode speech/data call or is in the call set up phase, TETRA Packet data is not allowed (accepted or initiated). This means that circuit mode speech/data may be set up, accepted or conducted while the MS is in STANDBY state. SDS may also be conducted while the MS is in the STANDBY state.

216


6.7 TETRA and WLAN Integration for Improving PacketData Transmission Capabilities2 Today, Public Safety and Disaster Recovery (PSDR) organizations rely extensively on Professional Mobile Radio (PMR) networks (mainly based on TETRA 1, 11 and Project 25 12, 13 specifications) in order to realize a mission critical communications environment as the one depicted in Fig. 6 (for illustration purposes, we depict a system based on TETRA technology). This environment is essentially a wide-area mobile radio network tailored to support the unique requirements of PSDR agencies, such as law enforcement, ambulance services, civil emergency management / disaster recovery, fire services, coast guard services, search and rescue services, government administration, etc. These agencies demand IP-based services (e.g. live video transmissions), which cannot be effectively supported by the TETRA packet data capabilities outlined in the previous sections, mainly due to bandwidth and QoS constrains. If we integrate Wireless Local Area Networks (WLANs) with TETRA networks, a considerable boost to packet data capabilities can be attained, given that WLANs can readily support packet transmissions with up to 100 Mbps rates in local areas. Evidently, such integration would be beneficial and we thus explore in this section the technical details of integrating WLANs with TETRA networks. Before doing so however, we present below the key components of PSDR networks, as they are typically deployed in practice today. This will provide enough background material for understanding later how these components can be integrated with WLANs. As illustrated in Fig. 6, the TETRA Switching and Management Infrastructure (SwMI) is the core component that comprises the necessary networking, switching, management and service provision elements of the system. It also provides digital narrowband radio services to a wide geographical area by means of a plurality of TETRA Base Stations (BS), deployed in strategic locations according to the overall radio coverage, traffic and availability requirements. The TETRA specifications define several types of air interface including the most commonly used Voice plus Data (V+D) air interface 1, the Packet Data Optimized (PDO) air interface 14 and the Direct Mode Operation (DMO) air interface 15 (see Fig. 6). The latter forms the means of enabling a fundamental mode of ad hoc networking, in which the radio users communicate directly to each other on a peerto-peer basis without the need of the TETRA infrastructure. This is particularly important in emergency situations that arise in locations outside 2 A modified version of this section has been published in IEEE Communications Magazine, see 9.

6.7 TETRA and WLAN Integration 217

the V+D coverage area. The V+D air interface operates in trunking mode, i.e. it allocates and releases the available radio resources dynamically and on demand basis. This way, the available radio spectrum can be efficiently shared across many different groups of users, or even across many different PSDR agencies.

Fig. 6. General configuration of a PSDR telecommunications system based on the TETRA technology.

As shown in Fig. 6, the SwMI also provides the essential interface components for linking the PSDR command and control centers with the portable, vehicular and airborne public safety users on the go, and for interoperating with conventional analogue (legacy) PMR networks, telephone networks, public or private data networks, or even with other TETRA SwMIs operated and administered by other organizations. Several IP-based data services are typically used in practice today, the most common being the Automatic Vehicular Location (AVL) service and the remote database query. In terms of services, the TETRA specifications developed by the European Telecommunications Standards Institute (ETSI) define a vast range of

218


voice, data and supplementary services, the most notable of which include dispatch services, group and private voice services, security services, telephone interconnect services, short data and status messaging and packet data services. All of these services are characterized by mission-critical performance figures such as fast call setup (typically, between 300-500 msec), fast message transmission, priority-based call handling, advanced encryption and authentication, etc. For a more comprehensive description of the TETRA services and technology the interested reader is referred to 16, 1, 10 -11. TETRA technology has enjoyed wide acceptance (especially in Europe) and is considered as one of the most mature and prominent technologies for the PSDR market as well as for the PMR/PAMR (Public Access Mobile Radio) markets. TETRA specifications are constantly being evolved by ETSI (see for example 17) and new features are being introduced to fulfill the growing and ever demanding PSDR requirements. With no doubt however, the mobile broadband technology could greatly enhance and complement the present and future TETRA networks and better achieve the advanced services envisioned in the next generation of PSDR communication systems; such as remote patient monitoring, 2-way real-time video, 3-D positioning and GIS, mobile robots, enhanced telemetry, etc. The required data rates for such advanced services plus the demand for enhanced mobility, improved ad-hoc functionality and international interoperability reach far beyond the scope of the current PSDR narrowband telecommunication systems and call for mobile broadband enhancements. To satisfy this need, the Project MESA (see www.projectmesa.org) was established between the ETSI and the Telecommunications Industry Association (TIA) in May 2000. The goal of this partnership project is to define a new and agile telecommunications system for the PSDR market by introducing a “system of systems” architecture, which would support a vast range of radio technologies (including wide, metropolitan and local area technologies as well as satellite technologies) and would integrate them into a robust advanced telecommunications network 18, 19. For fulfilling this goal, extensive scenario analysis has been used in order to identify, catalogue and document all these requirements into the so-called MESA Statement of Requirements (SoR) 20. This chapter aims at contributing to the MESA efforts toward the next generation of PSDR telecommunication systems by proposing a new scheme for seamlessly integrating TETRA and Wireless Local Area Networks (WLANs). The general concept is schematically illustrated in Fig. 7. The evolved PSDR system is composed of an IP-based TETRA SwMI that supports conventional TETRA Base Stations (BS) as well as WLAN hot


zones, deployed in strategic locations such as airports, government buildings, hospitals, etc. The WLAN hot zones connect to the TETRA SwMI via Interworking Function (IWF) elements and provide advanced TETRA services such as broadband data communications, true concurrent voice and data services, etc. The PSDR radio users are envisioned to carry dual mode terminals capable of supporting both conventional TETRA and WLAN radio interfaces.

Fig. 7. An evolved PSDR telecommunications system based on

WLAN/TETRA integration. In the rest of this chapter, we thoroughly describe a technical solution that can enable the integrated PSDR environment shown in Fig. 7. In particular, in the next section we provide a general overview of this technical solution and we present the prime benefits and new capabilities associated with it. Subsequently, we focus on the system architecture and we discuss its key functional elements and interfaces, as well as the protocol architecture. Moreover, we present several procedures and signaling scenarios, which help in demonstrating the operational details of the proposed solution. Finally, we conclude by summarizing the key features, capabilities and benefits that can be introduced in PSDR communication systems by applying the proposed WLAN/TETRA integration.

220


6.7.1 Integrated WLAN/TETRA System Overview The ongoing intense WLAN standardization and R&D activities worldwide, which target bit rates higher than 100 Mbps, enhanced security, improved mobility management, quality of service support and interworking with cellular networks, justify the fact that WLAN technology will play a key role in the wireless market. In light of this evolution and market trends, the incorporation of WLANs into TETRA networks is deemed as an interesting perspective capable of providing advanced new features to the PSDR communication systems. This forms the basic motivation for investigating potential means for integrating WLAN and TETRA networks and for proposing the technical solution presented in this chapter. The proposed system configuration is illustrated in the schematic diagram of Fig. 8. This system enables TETRA terminals to interface to the TETRA SwMI over a WLAN radio interface (instead of the conventional TETRA V+D radio interface) and over an IP access network. These terminals, referred to as TETRA-over-WLAN (ToW) terminals, can employ all typical TETRA services, including group calls, short data and status messaging, packet services, etc, but can also support a range of brand new services (this is further explained below). Thanks to the adaptation functions provided by the Interworking Function (IWF), the SwMI needs to provide no special functions to support the ToW terminals. In other words, both ToW terminals and conventional TETRA terminals are equally alike from the SwMI point of view. However, the characteristics of the WLAN radio interface enable ToW terminals to enjoy extended capabilities and new features, such as high-speed data services, inexpensive and simple provision of simultaneous voice and data services, improved voice quality, better call performance (reduced setup and voice transmission delays), simultaneous reception of many group calls, etc. The key component of the integrated system presented in Fig. 8 is the Interworking Function (IWF), which interfaces to the SwMI similarly to a TETRA Base Station (BS). The IWF interfaces also with one or more WLAN Access Gateways (WAGs) and uses IP multicasting technology to transfer control packet data units (PDUs) and voice packets to the ToW terminals. Mobile or fixed ToW terminals can access the typical services provided by the TETRA SwMI by means of a WLAN network interface and the appropriate software drivers and applications. For simplifying the implementation, ToW terminals re-use the majority of TETRA air interface protocols (see 1) and implement an adaptation layer for enabling their operation on top of the WLAN radio interface (see Protocol Architecture below). It is important to note that interfacing IWF with a TETRA SwMI


that is already built on IP multicast and Voice-over-IP (VoIP) technologies is relatively straightforward.

Fig. 8. General configuration of an integrated TETRA/WLAN public safety communications system. By integrating WLANs and TETRA networks as shown in Fig. 8 a list of beneficial features can be attained. Firstly, these hybrid networks can enjoy improved availability, reliability and network survivability. Indeed, by using dual-mode terminals and two different radio access technologies with overlapping coverage in strategic locations, the service availability is considerably improved in these locations. Also, the deployment of WLAN as an overlay radio access network leads to increased overall system survivability (i.e. provision of nominal radio services even during infrastructure component failures 21), which is of paramount importance in public safety and disaster recovery communication systems. Increased reliability is also doable especially in indoor areas, where the WLAN can provide better RF coverage and thus ensure less radio link failures and less dropped or missed calls. Secondly, a variety of revolutionary new features and capabilities can be enabled. For example:

222


Public safety officers and other mission critical users will be able to use their dual-mode TETRA/WLAN terminals in the field but also in the office environment, over their private WLAN. In strategic locations, such as airports, hotels, shopping centers, government buildings, etc, TETRA services could be provided over WLAN. This will enable better indoors coverage, increased capacity and novel new services in a cost effective way. In WLAN hot zones advanced services could be provided to the end user, such as: o Enhanced voice encoding schemes could be used for much improved voice quality. This is indeed possible because there is no need to restrict to codecs that operate within the 7.2 kbps capacity of the conventional TETRA speech bearers (TCH/S); o Broadband data services could be supported; o True simultaneous voice and data services could be easily provided (without the implementation complexity associated with TETRA Type-A terminals 1); o All control traffic normally transmitted on the TETRA common control channels could be received by WLAN radios while they have ongoing voice and/or data sessions active (this will be further explained later). This creates brand new capabilities such as easy ‘scanning’ of important group activities in the middle of a individual call or packet data session; o Already available IP applications could be readily exploited; o By means of IP multicast many group calls could be monitored simultaneously by a single radio user, etc. WLANs feature large air interface capacity and can support many simultaneous TETRA voice/data calls in an efficient and cost effective manner; Call setup delays and voice transmission delays could be considerably reduced due to the increased bit rates of WLAN; The WLAN network can easily be integrated with an IP-based network, which is envisioned to form the backbone of the next generation (MESA) public safety communication systems for meeting multiple different service requirements; A vast range of inexpensive WLAN-enabled subscriber devices can be used such as PCs, PDAs, or even cellular devices with WLAN capabilities, with the addition of suitable device drivers. Notably, dualmode WLAN/Cellular mobile terminals are quickly emerging in the

6.8 System Architecture 223

marketplace and these terminals could also support TETRA services over WLAN with moderate software enhancements. The sections below provide a comprehensive description of the proposed solution, including the system architecture, the definition of the functional elements, the protocol architecture, the packet structures, examples of signaling flows, etc.

6.8 System Architecture In this section we thoroughly discuss the architecture of the proposed system including the functional elements and the key interfaces of this architecture. Fig. 9 serves as the basis of the discussion.

Fig. 9. System architecture and traffic routing for various types of terminals

6.8.1 Architecture Elements and Interfaces AP: A typical WLAN Access Point, which interfaces with the WLAN terminals over any kind of WLAN interface. Without any loss of generality, in this chapter we assume that IEEE 802.11 WLAN technology is used.

224


WAG: The Wireless Access Gateway (which is physically a router or a combination of router and Ethernet Switch) controls one WLAN site. The WAG interfaces with one or more APs typically through an Ethernet 100BaseT medium. The WAG is responsible to create an Ft Tunnel with the IWF (see Fig. 9) when there are ToW terminals in its WLAN site and apply an appropriate routing enforcement policy (explained below). In addition, the WAG is responsible to release the Ft Tunnel when there are no ToW terminals in its WLAN site. The interface between the WAG and the IWF is referred to as Ft interface and is used essentially to manage the Ft Tunnel, which relays IP packets between the ToW terminals and the IWF through an IP network. Any possible tunneling scheme could be used, e.g. IP encapsulation, GRE, etc. Of course, in case where the IWF interconnects with a WAG over a leased line, tunneling can be eliminated. WLAN site: A WLAN site is defined as a geographical area wherein WLAN coverage is provided and it is controlled by a single WAG. A WLAN site is typically composed by one or more APs as shown in Fig. 9. IWF: The InterWorking Function (IWF) is a key functional element that interfaces with the TETRA SwMI over a proprietary interface and also interfaces with one or more WAGs (as shown in Fig. 8) over the Ft interface. From the SwMI point of view, the IWF corresponds to a TETRA Base Station that controls one location area. The location area controlled by a single IWF is called a WLAN Area and can be composed by one or more WLAN sites. The prime purpose of IWF is to hide from the SwMI the peculiarities of the WLANs and thus making it easier to integrate them with almost no SwMI changes. ToW terminal: A Tetra-over-WLAN (ToW) terminal is any WLAN terminal that can interface with the TETRA SwMI and employ TETRA services by means of the protocols and functions specified in this chapter. ToW terminals associate with the WLAN by using a special pre-defined Service Set Identifier (SSID), e.g. “TETRA”. By means of this special SSID the WLAN can differentiate between ToW terminals and non-ToW terminals and apply different routing and security policies. In particular, the WLAN implements a special routing enforcement policy for ToW terminals that tunnels packets from all ToW terminals to the IWF. This is schematically illustrated in Fig. 9, which shows how traffic from ToW terminals is tunneled to the IWF and how traffic from (non-ToW) terminals that use SSID=“PUBLIC” could be routed to another destination (e.g. the Internet or an Intranet). Apart from the different routing, the WLAN can also apply different access control policies for ToW and non-ToW terminals in order to meet the different security requirements of the


different terminal types. Typically, the ToW terminals will require strong authentication and encryption procedures, whereas the non-ToW terminals might have much more relaxed security requirements. Every ToW terminal implements a specific protocol architecture and procedures (specified below) in order to support TETRA services over WLAN. Physically, a ToW can be any kind of device with a WLAN interface, namely, a PC, laptop, PDA, dual-mode WLAN/TETRA terminal, etc. From the SwMI point of view, ToW terminals are like any other TETRA terminals. For example, each ToW terminal is assigned a regular TETRA Individual Short Subscriber Identity (ISSI), can initiate and participate in group calls, can receive/send Short Data messages, and in general can utilize all authorized services provided by the TETRA SwMI. In addition, each ToW terminal can communicate with other ToW terminals, with conventional TETRA terminals, with dispatchers, PSTN users, and other TETRA entities in accordance with its subscription profile in the SwMI. The ToW terminal communicates with the IWF over the Ut logical interface, which supports specific protocols and procedures. As discussed in the next section, a new protocol operates on this interface, namely, the Adaptation Layer. Also, this interface supports several other standard TETRA air-interface protocols, such as the LLC, SNDCP, etc 1. 6.8.2 Protocol Architecture Fig. 10 illustrates the proposed protocol architecture of a ToW terminal. The WLAN Physical and MAC layers are used for establishing broadband wireless connectivity with an AP. These layers are assumed here to comply with the IEEE 802.11 and 802.11g specifications, and with the QoS enhancements specified in IEEE 802.11e. They may also support the security features specified in the IEEE 802.11i specification. Connectivity with the IWF is provided with the IP layer, by means of its routing and addressing services while the UDP layer provides error detection and multiplexing services. On the User Plane, the Real Time Protocol (RTP) or the Compressed RTP protocol is used to transport ACELP encoded voice blocks between the ToW and the IWF. Although the typical TETRA voice codec is shown in Fig. 10, as pointed out before, any (wideband) voice codec could be used in order to enhance audio quality. On the Control Plane, all TETRA air interface protocols 1 are re-used except the TETRA MAC and Physical layer protocols, which are not applicable to WLAN radio access. The LLC layer supports both the Basic Link services and the Advanced Link services (see 1) and runs in the ToW and in the IWF.

226


Fig. 10. The protocol architecture of a ToW terminal.

A new layer is specified, namely, the Adaptation Layer, which provides the necessary adaptation functionality required to operate the TETRA air interface protocols over a WLAN. The Adaptation Layer is implemented in the ToW and in the IWF and provides services that include a subset of the services provided by the TETRA MAC layer. In particular, it supports TETRA-compliant encryption and addressing using TETRA Short Subscriber Identities (SSIs). As depicted in Fig. 10, the Adaptation Layer interfaces with the UDP layer through a Control-Plane Service Access Point (CP-SAP) and one or more User-Plane Service Access Points (UP-SAPs). The CP-SAP is always present and is used to carry control-plane traffic that is not associated with an ongoing call, i.e. control traffic normally transmitted on a TETRA Main Control Channel (MCCH) or Secondary Control Channel (SCCH) 1 for definition of these channels). A well-known (predefined) multicast IP address and Port number, referred to as MCCH-mcast and MCCH-port respectively, are specified here in order to transport such kind of traffic over the WLAN. On the other hand, user plane traffic is carried on a UP-SAP. At the beginning of a new call, a new UP-SAP instance is created to support that particular call. A UP-SAP instance uses a dynamically assigned multicast IP address and port number. As discussed below, this multicast IP address and port number are assigned by the IWF and are communicated


to the ToW in the packet that signals the start of the new call. A UP-SAP also supports control-plane traffic that is associated with an ongoing call (e.g. TETRA floor control messages). The Adaptation Layer is used to differentiate between the user-plane traffic and the call-associate control traffic on the same UP-SAP. The Adaptation Layer in the ToW analyzes every received packet and identifies (based on the associated SSI) if it should further be processed or get dropped. If it should further be processed, it decides if decryption should be applied and forwards it either to an LLC entity or to an RTP entity. More information about the Adaptation Layer is specified in the next section. 6.8.3 Packet Structure In Fig.11 we illustrate the general format of control-plane and user-plane packets proposed over the Ut interface (i.e. between the IWF and the ToW terminals). The control-plane packets carry normal TETRA LLC PDUs encapsulated into IP/UDP. The structure of the Adaptation Layer header, which is a key component of the proposed scheme, is described below. The structure of all other protocol fields (e.g. IP, UDP, RTP, LLC, CMCE, MLE, MM, SNDCP) complies with the structure in the corresponding protocol specification; for example see 1 for LLC, CMCE, MLE, MM, and SNDCP protocols and 22 for RTP protocol.

Fig.11.: Format of control-plane (a) and user-plane (b) packets transmitted between IWF and ToW.

228


The Adaptation Layer header is populated with critical information that is normally included in the TETRA MAC header. More specifically, the Adaptation Layer header includes a TETRA SSI and an Encryption Mode field, which indicates if the embedded LLC or RTP PDU is encrypted or not. In the downlink direction the TETRA SSI identifies the TETRA address of the packet recipient(s), whereas in the uplink direction, it identifies the TETRA address of the packet originator. In addition, the Adaptation Layer includes more information in packets that signal the origination of a new call (of any kind). In this case, the Adaptation Layer includes also the Multicast address and the Port Number that will be used to transport the voice packets of the upcoming call. Finally, the Adaptation Layer header includes an Info field that indicates if there is an LLC PDU or an RTP PDU encapsulated in the packet. 6.8.4 WLAN Association and TETRA Location Update Procedure In order to further explain the operational details for the proposed interworking scheme, we discuss in this section a simplified message sequence flow that can take place when a ToW terminal associates with the WLAN (by using SSID=TETRA) and then performs a Location Update in order to update the TETRA SwMI with its current whereabouts3. This happens either when the ToW terminal powers up or when it chooses to change radio access technology, i.e. to leave a conventional TETRA site and join a WLAN site. At point 1, the WAG creates a Ft tunnel with the predefined address of the IWF, if there is no such tunnel already in place. It also sets up its forwarding function so as to forward subsequent packets from the ToW to the IWF via the Ft tunnel. Next, the ToW initiates a DCHP procedure to attain IP configuration data, including an IP address. This IP address is typically assigned by the IWF, using either an internal or and external DCHP server.

Note that Fig. 12 represents only a typical example and does not aim at showing every single detail. 3


Fig. 12. Message sequence for WLAN association and TETRA location update. At point 2, the ToW enables its CP-SAP and starts receiving packets with destination IP and UDP port equal to the predefined MCCH-mcast and MCCH-port respectively. The values of MCCH-mcast and MCCHport are assumed to be pre-configured in the ToW. However, other means could also be developed for sending these parameters to the ToW, if necessary. After point 2, the Adaptation Layer in the ToW starts receiving and processing packets that include TETRA traffic normally transmitted on the MCCH channel such as D-MLE-SYSINFO, D-SETUP, D-SDS-DATA messages, etc (see 1 for details on these messages). After the ToW terminal identifies specific information pertaining to the TETRA SwMI (e.g. its unique network identity), it sends to the IWF a U-LOCATION-UPDATEDEMAND PDU to request the SwMI to update its location (point 3). This PDU is transmitted on the CP-SAP and therefore the destination IP address is equal to MCCH-mcast and the destination UDP port is equal to MCCHport. Finally, the SwMI accepts the location update request and responds with a D-LOCATION-UPDATE-ACCEPT PDU that is also transmitted by IWF on CP-SAP.

230


6.8.5 Group Call Initiation and Participation As a further signaling example, let’s assume that a conventional TETRA terminal with SSI=46 initiates a call to group 8388888, and that there are ToW terminals affiliated to this group. Fig. 13 illustrates the signaling flow that occurs for a ToW to receive this group call and to actively participate in it. The TETRA SwMI knows that there are subscriber affiliated to group 8388888 via the IWF and therefore will forward the call setup request to the IWF (point 1 in Fig. 13). Subsequently, the IWF sends to all WAGs under its controlling area a packet containing the corresponding D-SETUP PDU (message 2). The destination multicast address of this packet and the destination UDP port are the well-known MCCH-mcast and MCCH-port, respectively. The Adaptation Layer header in this packet indicates that the new call will use the IP multicast address g1.g2.g3.g4 and the UDP port Gp. All ToW terminals in the WLAN area of the IWF receive this packet, no matter if they are engaged in a call or not. ToW terminals affiliated to group 8388888 and willing to participate in this group call, will create a new UP-SAP instance and will bind it to the designated multicast address and UDP port (g1.g2.g3.g4 / Gp). Fig. 13 shows the ToW terminal with SSI=90, which receives this group call. After message 2, a series of IP multicast datagrams are transmitted, each one carrying a voice packet from the originator (see message 3 and message 4 in Fig. 13). All these datagrams indicate in the Adaptation Layer header that they carry encrypted RTP PDUs for the group 8388888. Message 5 is a call-associated control packet, carrying a D-TX Ceased PDU, indicating that the group call originator has ceased transmission. The Adaptation Layer in the ToW understands that this carries an LLC PDU (as opposed to an RTP PDU) and thus forwards it to the LLC layer (based on the Info field). After that, the considered ToW decides to take control of the group call and thus sends an uplink call-associated control packet (message 6), which carries a U-TX Request PDU that requests from the SwMI permission to transmit. In message 7, the SwMI grants transmit permission to ToW and, starting from message 8, the ToW transmits a series of user-plane packets that contain encrypted voice blocks encapsulated in RTP PDUs. Signaling flows corresponding to other call control procedures, such as for individual and/or telephone call establishment, are very similar to the signaling flows already described. Also, other procedures and corresponding signaling flows, e.g. for SDS and packet data transmission/reception can be easily developed by using the principles and the protocols described already. In addition, the typical TETRA authentication procedure can readily

6.9 Conclusions 231

be supported by exchanging the appropriate layer 3 messages between the ToW and the IWF, e.g. the D-AUTHENTICATION DEMAND, the U-/DAUTHENTICATION RESPONSE/RESULT (see 11). During this procedure, the Adaptation Layer in the ToW is responsible to run the appropriate security algorithms (as specified in 11) and create a Derived Ciphering Key (DCK). This key is subsequently used by the Adaptation Layer to encrypt and decrypt all LLC and RTP PDUs over the Ut interface. Support of security key management procedures (e.g. for Static and Common/Group Cipher Keys) is also feasible by exchanging the corresponding TETRA messages over the WLAN network.

Fig. 13. Message sequence for group call initiation and participation.

6.9 Conclusions Evidently, the narrowband packet data transmission capabilities of TETRA release 1 networks, as discussed above, is not good enough for meeting the demanding requirements of public safety agencies. For instance, supporting live video transmission from officers in the field to the command and control centers can be very problematic or even totally

232


infeasible when the TETRA PDCH is used. We showed however that when WLANs are integrated into wide-area TETRA networks, they can offer broadband packet data capabilities in strategic locations of the mission critical network. In this chapter, we extensively discussed a solution that enables such integration, and which allows TETRA terminals to interface to the TETRA SwMI over a broadband WLAN radio access network, instead of the conventional narrowband TETRA radio network. By means of this solution, the WLAN technology can be integrated into TETRA networks and enable new products and services for the public safety and disaster recovery (PSDR) market. The solution we presented is solely based on IP multicast and VoIP technologies and can thus fit ideally to the IP-based architecture envisioned by Project MESA. In this context, this solution can be deemed as a step forward in the evolution of PSDR communication systems. As a final note, we summarize below some of the key features and benefits of the discussed TETRA/WLAN integration solution.

It is based on IP multicast and VoIP technologies Can support seamless roaming between WLAN access and conventional TETRA access: Mobiles entering a WLAN area are like entering a new TETRA location area. They use the typical TETRA mobility management procedures to update the SwMI with their new whereabouts TETRA terminals can simultaneously participate in group call(s), data session(s) and also receive information normally sent on MCCH. This creates new capabilities not available on conventional TETRA radio systems TETRA technology is re-used as much as possible in order to minimize cost and development effort, and ensure seamless integration between WLAN and TETRA. The WLAN is primarily used as a new wireless access scheme for the transportation of TETRA control and user data PDUs. The impact on the SwMI is minimized. Indeed, the IWF can be considered as a special kind of TETRA Base Station, which can easily interface with the SwMI core. ToW terminals can be managed in the same way as any other conventional TETRA terminal for the SwMI point of view. Full compatibility between ToW terminals and conventional TETRA terminals is maintained.

References

233

References 1. ETSI EN 300 392-2 v2.6.1, “Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 2: Air Interface (AI),” May 2006. 2. Motorola Dimetra Release 5.1/5.2, Packet Data Service Programmers Guide. 3. Dimitrios I. Axiotis, Dimitrios Xenikos, “UDP Performance over TETRA,” to be published. 4. IETF RFC 768, “User Datagram Protocol,” August 1980, online at http://www.rfc-editor.org 5. IETF RFC 1661: “The Point-to-Point Protocol (PPP)” 6. IETF RFC 2865: “Remote Authentication Dial In User Service (RADIUS)” 7. IETF RFC 1994: “PPP Challenge Handshake Authentication Protocol (CHAP)” 8. ISO/IEC 8348: “Information technology - Open Systems Interconnection Network service definition” 9. A. K. Salkintzis, “Evolving Public Safety Communication System by Integrating WLAN and TETRA Networks,” IEEE Communications, vol. 44, Issue 1, pp. 38 – 46, Jan. 2006. 10. ETSI, EN 300 395-2 V1.3.1, “Terrestrial Trunked Radio (TETRA); Speech codec for full-rate traffic channel; Part 2 TETRA codec,” Jan. 2005. 11. ETSI, EN 300 392-7 v2.3.1, “Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 7: Security,” June 2006. 12. TIA/EIA-102.BAAA, “Project 25 FDMA Common Air Interface,” May 1998. 13. TIA, TSB102-A, “Project 25 System and Standards Definition,” Nov. 1995. 14. ETSI, ETS 300 393-2 second edition, “TETRA Packet Data Optimized (PDO) Air Interface,” Aug. 1999. 15. ETSI, EN 300 396-3, v1.2.1, “TETRA; Technical requirements for Direct Mode Operation (DMO); Part 3: Mobile Station to Mobile Station (MS-MS) Air Interface Protocol,” Aug. 2006. 16. ETSI, EN 300 392-1, v1.3.1, “Terrestrial Trunked Radio (TETRA); Voice plus Data (V+D); Part 1: General network design,” June 2005. 17. ETSI, TR 102 021-4, v1.3.1 “Terrestrial Trunked Radio (TETRA); User Requirement Specification TETRA Release 2; Part 4: Air Interface Enhancements,” May 2006. 18. M. Metcalf, “Project MESA: Advanced Mobile Broadband Communications for Public Safety Applications,” IEEE Int. Symposium on Personal, Indoor and Mobile Radio Communications, 2003. 19. S. Ring, “Mobile Digital Communication for Public Safety, Law Enforcement and Non-Tactical Military,” available at http://www.etsi.org/T_news/ Documents/TETRA-TETRA2-MESA.pdf, Jan. 2001. 20. Project MESA, technical specification TS 70.001 V3.1.2, “Statement of Requirements,” Jan. 2005.

234


21. U. Varshney, A.P. Snow and A.D. Malloy, “Measuring the reliability and survivability of infrastructure-oriented wireless networks,” 26th Annual IEEE Conference on Local Computer Networks (LCN’01), Nov. 2001. 22. H. Schulzrinne et al., “RTP: A Transport Protocol for Real-Time Applications”, RFC 1889, 1996.

7 TETRA as a Building block to WMNs

Pau Plans, Carles Gomez, Josep Lluis Ferrer, and Josep Paradells

7.1 Introduction Communications solutions for emergency services are nowadays based on two clear differentiated trends: on one side, the digital Professional/Private Mobile Radio (PMR) technology with trunking facilities and, on the other one, the well known Wireless Mesh Networks (WMNs) 2. PMR systems, which have been continuously adapted to the users needs from their appearance, are currently fulfilling the most essential requirements from the users. WMNs show also promise for emergency applications because of their suitability for these environments 1. These networks may operate without infrastructure and may be deployed with no previous planning. Besides these two well defined tendencies, there are also other proposed solutions which are based on existing systems or combinations of them. As will be shown later, no solution offers the same level of reliability than that offered by current PMR systems. However, some aspects should be improved. For example, present PMR systems are only able to offer narrowband services (voice and short messaging) and they are based on the already in-place infrastructure. For networks devoted to emergency applications this can result in a serious compromise with cost implications. Most of the time the network is not used (because it is an emergency network) and when necessary, it should be available and provide enough capacity. Emergency networks are not dimensioned for daily operation, they should be able to offer extra capacity to overcome the emergency situation. Predicting their size is an extremely

236


difficult goal. The September 11th tragedy in New York demonstrates the underdimensiong of current emergency communication systems. WMNs can be suitable for emergency cases since they are based on a totally different approach from that of PMR systems. WMNs do not require infrastructure and the reliability and the capacity of the network improves as the number of users increases. Unfortunately, present mesh solutions are based on different technology from that used in PMR. Thus, interworking between both networks is an issue. The possibility to build a mesh network based on TETRA seems a promising option to get the best of both worlds. In this chapter we analyze possible extensions of the TErrestrial Trunked RAdio (TETRA) 3 system with the purpose of building a mesh network 4. The main objective is to provide the services, already offered by PMR systems, everywhere and at any moment, taking then profit of the natural advantages that a mesh network presents. We evaluate extensions which make use of already available functionality of TETRA, such as client, relay and gateway functions, in order to minimize the changes that should be made to the standard and make the adaptation as simple as possible. Although our proposal is focused on TETRA, it also may be modified to other similar PMR digital systems like the Association of Public-safety Communications Officials (APCO 25). Most of the communications solutions for emergency services currently being deployed around the world are based on TETRA or APCO 25. Both are digital narrowband technologies. TETRA, which is today a mature technology, enjoys a wide acceptance around the world. Many countries, not only in Europe (where TETRA was born), are today deploying TETRA networks. Although TETRA covers the user requirements, it needs, as any other promising standard, to go forward as time progresses. The European Telecommunications Standardization Institute (ETSI) is constantly evolving TETRA specifications to better satisfy the emerging user requirements and therefore ensure longevity. TETRA release 2 suite of standards is currently being defined. Some of its main purposes are to extend the current range of coverage, to introduce an enhanced voice codec, and to enhance data services using different technologies while maintaining compatibility with TETRA release 1. The evolution of TETRA towards the release 2 is a future proof investment which maintains

7.1 Introduction 237

the competitiveness with other wireless technologies in the PMR and Public Access Mobile Radio (PAMR) standards field. APCO 25 is the other PMR system in which most public safety systems of the world are based. It is standardized by the Telecommunication Industry Association (TIA) and the Electronic Industries Alliance (EIA). APCO 25, similarly to the evolution of TETRA, is going towards an enhanced release in order to support more bandwidth demanding services and applications such as multimedia messaging, high-resolution image transfer, video… TIA 902, 56 which is an overlay technology aimed to offer high data rates, has been developed with this purpose. WMNs are one of the main focuses of the research community, especially in the emergency networks sector. WMNs constitute a promising technology as a solution for public safety and first-responder networks 1. They are usually based on IEEE 802.11 standards, which make use of higher frequencies involving more available bandwidth than those offered by current PMR systems. This radio technology, however, is not more appropriate than that of PMR digital systems. Some of the main reasons are: it presents limitations in front of high users’ mobility, it works in an unlicensed ISM band which may be affected by interference and the high frequencies used experiment short distance coverage and low building penetration. The problems due to the use of an ISM band may disappear with the allocation of an emergency band at 4,9 GHz, such as the one done by the Federal Communication Commission (FCC) recently, but propagation losses remain as the main problem. Although, in this band, power can be boosted to allow better building penetration, coverage restrictions may appear due to the channel delay spread, especially in the case of outdoor use 7. Another proposal for emergency networks that has been discussed so far is the usage of public commercial networks for emergency communications 8. Nowadays, however, public networks should still improve many aspects in order to reach the reliability that native emergency solutions present. Recognizing the possibility of using public public cellular system for PMR servives, the Open Mobile Alliance (OMA) presents the first open standard for industry aimed to offer Push To Talk (PTT) over Cellular (PoC). This effort however dos not currently address some of the users’ requirements, especially in the emergencies sector, where overlaid solutions may be applied 9.

238


Besides these specific solutions, other approaches consider the usage of more than one technology in one system. The Mobility for Emergency and Safety Applications (MESA) project 10 is an international partnership that was initiated between TIA and ETSI in the 2000. The goal of MESA is to produce technical specifications for digital mobile broadband technology, originally intended for the public safety and disaster response sector. MESA introduces a “system of systems” architecture. The SAFECOM program 11, a communication program of the U.S. Department of Homeland Security’s Office for Interoperability and Compatibility (OIC), supports the project MESA. The Wireless Deployable Network System (WIDENS) 12 project has contributed to the MESA standardization process with a development of an ad-hoc rapidly deployable broadband network demonstrator and system concept for public safety. The goal is to provide high data rate hotspots in order to enhance the public safety, emergency and disaster applications, being complementary and interoperable with existing infrastructures (TETRA and Tetrapol) at the same time 13. Project WIDENS includes a solution called Cross Standard System (CSS) for future public safety and emergency communications 14 which attempts to solve one of the main problems of public safety communications: the interoperability between different communications systems. The CSS is based on the support of several air interfaces in order to make use of different standard communication systems (WLAN, public networks, PMR standards, etc). This is another example where the need of interoperability is highlighted. Another approach which combines different technologies is described in 15. This proposal suggests the integration of TETRA and WLAN. A TETRA terminal may interface with TETRA infrastructure over WLAN access network. It requires the usage of ToW (TETRA over WLAN) dual terminals. The TETRA network, in this case, treats a WLAN area (which can be composed of one or more WLAN sites) as a TETRA base station. The WLAN peculiarities are hidden from the TETRA network by means of a specific gateway. As a matter of conclusion on the topic of alternative systems to PMR, it can be observed that there are other technologies that can complement PMR systems such cellular public systems for improving the number of user or Wireless LAN mesh networks for broadband services, but none of


them can replace PMR for emergency services. For group communications using voice PMR, and in our case TETRA, has no competitor. PMR systems account with a variety of users, covering many fields, namely: public safety and public security, transportation, utilities, government, military, PAMR (Public Access Mobile Radio), commercial, etc. Each kind of user has specific requirements. 7.1.1 Requirements In this section we enumerate the most important requirements that users demand, which can be translated to network requirements. Depending on the network technology used these requirements may be satisfied more easily. It is clear that as the requirements are relaxed it is more probable to find different technologies to satisfy them. What would be clear also is that the usage of a wireless mesh network based on TETRA would relax the amount of infrastructure needed to satisfy the requirements. Reliability

Users demand availability of the services. The network has to have enough redundancy to ensure continuity of the service even in case of network fail. There are different approaches to increase reliability of the service. Key elements of the network should be redundant in order to guarantee the service even in case of failure. In case a problem occurs and the redundancy does not exist or not enough the system should keep working delivering services at least in a local manner. To protect the system against a base station failure the coverage should be overlapped allowing the rest of base stations to attend the users of the faulty base station. It can be seen that reliability is a manner of redundant equipment or coverage. In terms of efficiency it is very poor, it means equipment that is in place and only used on case of element failure. The possibility to deploy an ad-hoc network when needed relaxes the amount of permanent redundant equipment. Coverage

Most of the usages of PMR networks require very good levels of geographic coverage. An emergency can occur at any place and any time and the network should be able to attend the communication. Assuring a full coverage on the territory results very expensive if not impossible to achieve. Furthermore, PMR systems may offer indoor coverage by means of increasing the power transmission level. But even with this configuration

240


indoors coverage is not achieved at any location. Mesh networking can offer coverage extension wherever it is needed and relax the infrastructure coverage. This advantage is very relevant in general, but is the only way to offer full coverage indoors. Speech and data transmission capability

Voice communication is a hard requirement for PMR system, but short messaging is becoming important. In particular automatic location that allows a continuous tracking of mobile terminal, make usage of short message services as a way to deliver the location provided by a GPS. There is a requirement for data transmission services in particular broadband data transfer such image transfer or video streaming, but such requirement is not feasible with present equipment. The usage of a parallel network such a WMN based on WLAN technology or the improvement to TETRA Release 2, when available, will offer the solution to the requirement. Speech communication is done mainly in half duplex. This possibility allows reducing the quality of service requirements needed for a full duplex communication. Also as one channel per conversation the number of users sharing a channel can be increased. Group communications

Users require different communication types, in particular group communication is the one most used. With this mode it is simple to exchange information within the group and it is very efficient in terms of resources used. There is no maximum size for the group as the number of users will depend on its activity. As the communication is activated as a push to talk fashion it is possible to allocate more than one group per channel. Group communication is simple to implement when all the participants are in the same radio coverage. When terminals are further away the network should allow this communication using techniques close to multicast routing. Geographical span

Communication can range from small areas, users located close one to each other (several meters), to a large area, for example the whole country. Traditional PMR systems use two technologies to attend this geographical span. For small area it is available the direct mode which supports the communication without infrastructure. When the communication area extends further it is needed to use the infrastructure. The area covered by direct mode depends on several factor such transmission power, frequency


band or scenario of usage (indoors, outdoors urban area, outdoors rural area,…), but can range from meters to several kilometers. The introduction of mesh networks ideas in PMR systems can extend the range of direct mode communication towards the one provided by infrastructure based solutions. Capacity

Public safety systems, in particular those used in emergency situations are difficult to dimension. In general the daily operation of the users it is known and the resources needed can be estimated considering the quality of service requirements from the users. In case of emergency the usage of the system changes: the area of usage, the number of users and the usage pattern is modified. Planning a system for an emergency results very difficult since there are not statistical models. The common approach is a slight over dimensioning of the system. In a real emergency this extra capacity will not be enough. This over dimensioning is expensive in terms of resources since implies infrastructure that is not used. Traditional PMR systems have some helps in order to cope against the lack of resources in case of an emergency. For example it is available a mobile base station that can be connected thought satellite link to the rest of the infrastructure. But these are patches to the lack of flexibility of the system. Mesh networks are by definition flexible and can be accommodated to the user demands. In case of emergency network can be deployed to attend the emergency, and once it is over network can be moved to another location. In fact mesh networks can be built by the users themselves, so when users concentrate they create the network.

Performance requirements of a communication

The performance requirements related to a communication mainly include: call set-up time, end-to-end delay and changeover time. - Call set-up time. It is the time spent for establishing the call. Different users may have different requirements for this performance parameter. Emergency services for instance requires fast call set-up because orders or information are usually urgent. Current public systems may last approximately seconds for call set-up which may not be acceptable in some contexts. The PTT operation, where calls can be established one-way, allows a fast call set-up. PMR systems may offer call set-up times of less than 300 ms. Note that this time includes the end-to-end delay (see next bullet). In

242


the PoC technology, the OMA defines a set of requirements that should be accomplished in order to offer certain Quality of Experience (QoE) 16. The call set-up requirement defined by the OMA is classified into two time values: -

-

Right-to-speak (RTS) is the time that an initiating user has to wait to be ready to begin transmissions. This time depends on the answer mode adopted for the destination user: automatic or manual. If automatic answer mode is set, the source user may have granted the right before the destination be reached (It is not necessary to contact with the destination because its confirmation is not needed). If manual answer mode is used, the destination user has to accept the invitation before the RTS can be given to the source. The time lasted between the initiation of the session by the source node (intention to transmit) and the moment when RTS is received (source may begin to transmit) should typically be less than 2 s when automatic answer mode is applied. If manual mode is used, RTS should typically be received in less than 1.6s after the destination part accepts the session invitations. Start-to-Speak (StS) is the time interval between the instant in which a participant requests permission to talk to the moment when it receives the permission. If requests are not queued, the StS should typically be less than 1.6 s. This is also the time within which a user should receive an indication of rejection or queuing of the request if applies.

Current PMR systems do not include the RTS time in the call set-up time. They only take into account the StS plus the end-to-end delay. - End-to-end delay. End-to-end channel delay is also defined as the mouth-to-hear delay. It is the duration between the instant in which the source user speaks and the moment in which it is heard by the destination. The PoC defines that it should typically be less than 1.6 s for general talkbursts. For the first talk-bursts, it should take no more than 4 s (note that this is the concept of call set-up when RTS and StS delays are not included because operations are no needed). Most users may be satisfied with these figures but others, e.g. the emergency ones, may require shorter times. - Changeover time, also called Turnaround Time (TaT), refers to the time a changeover between users occurs. i.e., the time between the instant in which the user stops talking and releases the channel and the instant in which another user begins to speak. According to the PoC requirements, in order to allow fluent communications, TaT should be no longer than 4s when users reply immediately (in 1-2 seconds).

7.1 Introduction 243 Call priorities

Emergency calls must have higher priority than other calls. They must even be able to throw out the lowest priority calls in case of network congestion. This requirement tries to guarantee availability of resources for emergency calls in case of network congestion. This requirement is mostly demanded by emergency services. Users in other contexts do not usually require call priorities. Voice quality

Voice quality takes special relevance in emergency situations because a high level of noise is usually present. It is then important to eliminate this background noise in order to make an understandable communication possible. Some PMR systems present high voice quality. Voice quality level required by PoC is not as strong as the one offered by PMR solutions, PoC requires a MOS >= 3 at BER

TErrestrial Trunked RAdio - TETRA: A Global Security Tool (Signals and Communication Technology)

Continuous-Time Signals (Signals and Communication Technology)

Communication Acoustics (Signals and Communication Technology)

Radio Wave Propagation for Telecommunication Applications (Signals and Communication Technology)

My personal Adaptive Global NET (MAGNET) (Signals and Communication Technology)

Speech Enhancement (Signals and Communication Technology)

Continuous-Time Systems (Signals and Communication Technology)

Blind Speech Separation (Signals and Communication Technology)

Speech Dereverberation (Signals and Communication Technology)

Three-Dimensional Television: Capture, Transmission, Display (Signals and Communication Technology) (Signals and Communication Technology)

Digital Communication: Principles and System Modelling (Signals and Communication Technology)

Fast Fourier Transform - Algorithms and Applications (Signals and Communication Technology)

Wireless Sensor Networks and Applications (Signals and Communication Technology)

Foundations and Applications of Sensor Management (Signals and Communication Technology)

Three-Dimensional Imaging, Visualization, and Display (Signals and Communication Technology)

Satellite Communications and Navigation Systems (Signals and Communication Technology)

Digital Video and Audio Broadcasting Technology: A Practical Engineering Guide (Signals and Communication Technology)

Imaging for Forensics and Security: From Theory to Practice (Signals and Communication Technology)

Coping with Interference in Wireless Networks (Signals and Communication Technology)

Acoustic MIMO Signal Processing (2006) (Signals and Communication Technology)

Voice over IP in Wireless Heterogeneous Networks: Signaling, Mobility and Security (Signals and Communication Technology)

Ad-Hoc Networking Towards Seamless Communications (Signals and Communication Technology)

Digital Video and Audio Broadcasting Technology: A Practical Engineering Guide (Signals and Communication Technology)

Remote Sensing with Imaging Radar (Signals and Communication Technology)

Sensor Networks: Where Theory Meets Practice (Signals and Communication Technology)

Satellite and Terrestrial Radio Positioning Techniques: A signal processing perspective

Technology and Scholarly Communication

Multimodal User Interfaces: From Signals to Interaction (Signals and Communication Technology)

Terrestrial Global Productivity (Physiological Ecology)

Terrorism and Global Security

Mobile Communication and Society: A Global Perspective

TErrestrial Trunked RAdio - TETRA: A Global Security Tool (Signals and Communication Technology)

Continuous-Time Signals (Signals and Communication Technology)

Communication Acoustics (Signals and Communication Technology)

Radio Wave Propagation for Telecommunication Applications (Signals and Communication Technology)

My personal Adaptive Global NET (MAGNET) (Signals and Communication Technology)

Speech Enhancement (Signals and Communication Technology)

Continuous-Time Systems (Signals and Communication Technology)

Blind Speech Separation (Signals and Communication Technology)

Speech Dereverberation (Signals and Communication Technology)

Three-Dimensional Television: Capture, Transmission, Display (Signals and Communication Technology) (Signals and Communication Technology)

Digital Communication: Principles and System Modelling (Signals and Communication Technology)

Fast Fourier Transform - Algorithms and Applications (Signals and Communication Technology)

Wireless Sensor Networks and Applications (Signals and Communication Technology)

Foundations and Applications of Sensor Management (Signals and Communication Technology)

Three-Dimensional Imaging, Visualization, and Display (Signals and Communication Technology)

Satellite Communications and Navigation Systems (Signals and Communication Technology)

Digital Video and Audio Broadcasting Technology: A Practical Engineering Guide (Signals and Communication Technology)

Imaging for Forensics and Security: From Theory to Practice (Signals and Communication Technology)

Coping with Interference in Wireless Networks (Signals and Communication Technology)

Acoustic MIMO Signal Processing (2006) (Signals and Communication Technology)

Voice over IP in Wireless Heterogeneous Networks: Signaling, Mobility and Security (Signals and Communication Technology)

Ad-Hoc Networking Towards Seamless Communications (Signals and Communication Technology)

Digital Video and Audio Broadcasting Technology: A Practical Engineering Guide (Signals and Communication Technology)

Remote Sensing with Imaging Radar (Signals and Communication Technology)

Sensor Networks: Where Theory Meets Practice (Signals and Communication Technology)

Satellite and Terrestrial Radio Positioning Techniques: A signal processing perspective

Technology and Scholarly Communication

Multimodal User Interfaces: From Signals to Interaction (Signals and Communication Technology)

Terrestrial Global Productivity (Physiological Ecology)

Terrorism and Global Security

Mobile Communication and Society: A Global Perspective

Recommend Documents