Wireless Lan Standards and Applications (Artech House Telecommunications Library)

Wireless LAN Standards and Applications For a listing of recent titles in the Artech House Mobile Communications Seri...

Author: Asuncion Santamaria | Francisco J. Lopez-Hernandez

82 downloads 1521 Views 1MB 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

Wireless LAN Standards and Applications

For a listing of recent titles in the Artech House Mobile Communications Series, turn to the back of this book.

Wireless LAN Standards and Applications Asunción Santamaría Francisco J. López-Hernández Editors

Artech House Boston London www.artechhouse.com

Library of Congress Cataloging-in-Publication Data Wireless LAN standards and applications / Asunción Santamaría, Francisco López-Hernández, editors. p. cm. (Artech House mobile communications library) Includes bibliographical references and index. ISBN 0-89006-943-3 (alk. paper) 1. Local area networks (Computer networks). 2. Wireless communication systems. I. Santamaría, A. (Asunción). II. López-Hernández, F. J. III. Series. TK5105.7 .W575 2001 004.68dc21 2001022207

British Library Cataloguing in Publication Data Wireless LAN standards and applications.(Artech House mobile communications library) 1. Local area networks (Computer networks) 2. Wireless communication systems I. Santamaría, Asunción II. López-Hernández, F. J. (Francisco J.) 621.3981 ISBN

1-58053-428-7

Cover design by Igor Valdman

© 2001 ARTECH HOUSE, INC. 685 Canton Street Norwood, MA 02062 All rights reserved. Printed and bound in the United States of America. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher. All terms mentioned in this book that are known to be trademarks or service marks have been appropriately capitalized. Artech House cannot attest to the accuracy of this information. Use of a term in this book should not be regarded as affecting the validity of any trademark or service mark. International Standard Book Number: 0-89006-943-3 Library of Congress Catalog Card Number: 2001022207 10 9 8 7 6 5 4 3 2 1

To Marcos, Daniel, Lucas, Luisa, Javier, Jaime, and Raquel, and to all the people who encouraged us to do this book.

.

Contents Preface

xi

Acknowledgments

xi

Introduction

1

1.1 1.2 1.3

1 2 3

Introduction to wireless local area networks (LANs) The need for standardization Future trends

References Selected Bibliography

7 8

The IrDA Standard

9

2.1 Introduction to the IrDA General description 2.1.1 The standard protocol stack 2.2 Physical layer (SIR) 2.2.1 General description of the physical layer specification

9 11 13 13

2.2.2

Specifications for signals at bit rates from 2.4 Kbps to 115.2 Kbps

14

2.2.3

Specifications for signals at bit rates of 0.576 Mbps and 1.152 Mbps

15

2.2.4

Specifications for signals at bit rates of 4 Mbps

17

2.2.5 Optical interface characteristics 2.3 Serial infrared link access protocol (IrLAP) 2.3.1 Services provided by IrLAP: Connectionless services

21 22 24

2.3.2

Services provided by IrLAP: Connection-oriented services

24

2.3.3

Configurations and operating characteristics

25

2.3.4

IrLAP frame structure

26

2.3.5

IrLAP frame types

26

vii

viii


2.4 IRDA link management protocol (IrLMP) 2.4.1 Link management multiplexor (LM-MUX)

28 28

2.4.2 Information access service (IAS) 2.5 IRDA transport protocol: TinyTP 2.5.1 TinyTP frames format

33 35 36

2.5.2 Flow control 2.6 LAN access extensions for link management protocol: IrLAN 2.6.1 IrLAN general description

36 38 38

2.6.2

Access methods

39

2.6.3

Frames size and format

41

References

43

The IEEE 802.11 Standard

45

3.1 Introduction to IEEE 802.11: General description 3.2 Medium access control (MAC) for the IEEE 802.11 wireless LANs (WLANs) 3.2.1 Expected features of a WLAN MAC protocol

45 47 48

3.2.2

The structure of the IEEE standard MAC protocol

54

3.2.3

Comparison with the MAC protocol of other WLANs: HIPERLAN

65

3.2.4 Conclusions 3.3 Physical layer for IEEE 802.11 wireless LANs: Radio systems 3.3.1 Introduction

69 69

3.3.2

Spread spectrum techniques

71

3.3.3

Frequency hopping techniques

72

3.3.4

Direct sequence systems

75

3.3.5

IEEE 802.11 frequency hopping physical layer

83

3.3.6

IEEE 802.11 direct sequence physical layer

89

3.3.7 Comparison of the FHSS and DSSS physical layers 3.4 Physical layer for IEEE 802.11 wireless LANS: Infrared systems 3.4.1 Description

93

3.4.2

The physical layer convergence procedure (IR-PLCP)

69

96 96 97

3.4.3 The IR physical medium sublayer (IR-PMD) 3.5 Conclusions and applications

99 104

References

105

Contents

ix

The HIPERLAN Standard

109

4.1 Introduction: Terminology 4.2 Physical layer (PHY) 4.2.1 Introduction

109 113 113

4.2.2

Transmission characteristics

114

4.2.3

Data bursts

116

4.2.4

Channel access bursts

118

4.2.5

Receiver characteristics

118

4.2.6

Compatibility between transmitter and receiver classes

119

4.2.7 Establishing a defer threshold 4.3 HIPERLAN channel access control (CAC) 4.3.1 Generalities

119 121 121

4.3.2

HIPERLAN CAC protocol data units (HCPDUs)

122

4.3.3

Channel access

124

4.3.4

Channel access in the free channel condition

126

4.3.5

Channel access in the synchronized channel condition

127

4.3.6 Hidden node detection and operation 4.4 HIPERLAN medium access control (MAC) 4.4.1 HIPERLAN MAC functions

130 131 131

4.4.2

HIPERLAN differentiation and addressing

132

4.4.3

Data encryption

133

4.4.4

Power-saving function

133

4.4.5

MAC information databases

135

4.4.6

Priorities and traffic lifetime

137

4.4.7

Types of HMPDUs

139

4.4.8

Look-up function: LR-HMPDU and LC-HMPDU

139

4.4.9

IP-HMPDU and GP-HMPDU

140

4.4.10

DT-HMPDU

140

4.4.11

TC-HMPDU and HO-HMPDU

141

4.4.12 Routing functions and information maintenance 4.5 Conclusions on HIPERLAN type 1 4.6 Future BRAN standards

142 146 147

References Selected Bibliography

148 149

Application Scenarios 5.1

Introduction

151 151

x


5.2 The application scenarios 5.2.1 Public buildings

152 152

5.2.2

Business environment

158

5.2.3

Domestic buildings (the home)

159

5.2.4 Industrial sector 5.3 Wireless LAN technologies and products 5.3.1 The RF market

160 161 161

5.3.2 The IR market 5.4 Conclusions

168 176

References Selected Bibliography Appendix 5A: IrDA membership

177 177 178

Upcoming Standards and Future Trends 6.1 Introduction: Is the future wireless? 6.2 The evolution of HIPERLAN 6.3 The evolution of IEEE 802.11 6.4 Forthcoming IR standards 6.4.1 IrDA new techniques 6.4.2

Interconnection for wireless networks

183 183 186 188 190 190 195

6.4.3 New techniques for diffuse links: Spread spectrum 6.5 Other RF standards: DECT, Bluetooth, WATM, HomeRF, etc. 6.5.1 Introduction

197 200 200

6.5.2

Digital enhanced cordless telecommunications (DECT)

201

6.5.3

Bluetooth

203

6.5.4

Wireless ATM

206

6.5.5 HomeRF 6.6 Conclusions

207 209

References

209

Glossary

213

About the Author

223

Index

225

Preface This book provides an overview of wireless systems standards. Moreover, to help readers stay current on these rapidly changing technologies, the book points readers to updated information on the Internet, providing the Web addresses of the main regulation and standardization organizations and listing the Web sites of the technological alliances driving the evolution of wireless technology. The book pays special attention to a handful of new environments in which wireless systems are gaining importance. In addition, the book highlights the prominent role that wireless technologies will play in the near future of home, office, transportation, and personal communications. With the most important information technology (IT) companies supporting the wireless systems field, its success is apparent; a new wireless world is becoming real.

Acknowledgments The authors offer special thanks to Luis Santamaría for his collaboration on this book.

xi

.

CHAPTER

1

Contents 1.1 Introduction to wireless local area networks 1.2 The need for standardization 1.3

Introduction A. Santamaría and F. J. López-Hernández

Future trends

1.1 Introduction to wireless local area networks (LANs) As we enter the 21st century, IT and communications are leading a revolution in the way we live. The number of cellular telephone subscribers and people accessing the Internet, the growth of electronic business, and the abundance of companies on the Web are leading the evolution of information technologies. Access to information has become a necessity. The information available on the Internet and the easy access to this information allows professionals of many different fields to use tools that would have been unavailable without this communication facility. Competitive, up-to-date companies are offering their services and products on the Web. Members of the research community are presenting their ideas and results via

1

2


the Web. The Internet has become an indispensable consultation tool as well as an open window in which people can showcase their information to the world. Taking a look at the job opportunities that now appear in newspapers, journals, and Internet job forums, we find that the demand for IT professionals is enormous. There is fierce competition among companies for these professionals. In fact, the development of systems and applications is being held back because of a lack of suitably trained professionals. In the IT field, there are two types of companies: the longestablished, well-known firms and the young start-ups. Those in the first category, such as Cisco Systems, Hewlett-Packard, and Lucent, are devoted to the development and commercialization of existing products. On the other hand, the start-ups provide innovative ideas, some of which succeed and some that fail. It is interesting to note that the research effort in wireless LAN (WLAN) systems that started some 20 years ago has allowed for the development of new ideas that have yielded applications and systems. Moreover, the established companies are actively involved in this booming market. Chapter 5 presents a comprehensive study of these new applications and systems and the companies involved in promoting them.

1.2

The need for standardization

With so many new wireless systems and applications from so many suppliers in so many countries, the need for standardization is essential. As a result, cooperation between wireless manufacturer and user interest groups has given rise to the creation of open associations to develop standards. One of the main standardization groups, the Infrared Data Association (IrDA), focuses on the development of infrared (IR) wireless systems, IEEE 802.11, high-performance radio LAN (HIPERLAN), Bluetooth, and home radio frequency (RF) for the evolution of different applications of RF wireless systems. Chapters 2, 3, and 4 detail the main characteristics of these standards, while Chapter 6 reflects on the future and the potential of these standards and describes the new and emerging standards in this field.

Introduction

1.3

3

Future trends

Wireless systems are evolving toward the development of broadband applications, including multimedia services. The aim of todays research is to achieve high-bandwidth wireless systems with performance comparable to that of high-bandwidth fixed networks. Focusing on fixed systems, the bandwidth available for the traditional public-switched telephone network (PSTN) can be noticeably increased by using new technologies such as asynchronous transfer mode (ATM), the integrated services digital network (ISDN), broadband ISDN (B-ISDN), and the different digital subscriber loop (DSL) possibilitiesasymmetric DSL (ADSL), high-speed DSL (HDSL), symmetric DSL (SDSL), and very high-speed DSL (VDSL). The development of wireless systems is subject to several technical limitations, such as transfer capacity, quality, and range. The environments in which wireless and mobile systems can be used are classified into five categories, depending on the range of cover [1]: ω

Home-cell environments for in-house applications;

ω

Picocell environments for in-building systems and applications;

ω

Microcell environments for applications covering urban areas;

ω

Macrocell environments for applications covering suburban areas;

ω

The global environment for applications using satellite-based systems (see Figure 1.1).

Systems and applications for the home-cell and picocell environments have evolved over the last decade, and today there are standards and commercial equipment that allow for the installation of wireless systems using RF and IR technologies. This book deals with these kinds of systems. There are three main standards that assure the compatibility of wireless communication systems for indoor applications. They are IEEE 802.11, HIPERLAN, and IrDA. IEEE 802.11 is a standard that covers RF and IR technologies for WLAN applications. Meanwhile, HIPERLAN deals with RF systems, and IrDA concentrates on IR systems. However, all three standards share a common application environment. They are all used for local applicationseither for the expansion of existing wired networks, the installation of new wireless networks, or special applications different from the

4


Global environment

Picocell Macrocell

Microcell Home-cell

Figure 1.1

Environment classification for wireless systems applications.

standard LAN (voice transmission in multilanguage rooms, for example). Wireless indoor systems are the ideal complement to buildings that have been designed with a structured cabled infrastructure, because they allow users to expand the cabled network, reallocate terminal equipment, add new segments, and install temporary working groups in an easy, cheap, and fast way. Special attention must be paid to the development of the new emerging standards, such as Bluetooth, wireless ATM (WATM), and home RF. It is expected that they will undergo a great deal of development during the next few years. The wireless systems for microcell and macrocell environments are now enjoying an enormous boom. Thanks to these systems, todays connections are person-to-person instead of traditional point-to-point connections. The spectacular growth in cellular telephony and the development of new services for mobile phones related to the Internet have created the need for a new generation of mobile systems. In examining the evolution of cellular telephone networks, it is possible to see tremendous change since the introduction of the analog-based first generation in the mid 1980s. The second-generation (2G) wireless networks, which use digital modulation, have afforded the implementation of integrated speech and data services. Some examples of 2G wireless networks are the global system for mobile communications (GSM), digital enhanced cordless telecommunications (DECT) in Europe, time division multiple

Introduction

5

access (TDMA) IS-54/IS-136 and code division multiple access (CDMA) IS-95 in the United States, and personal digital cellular (PDC) in Japan [2]. The third generation (3G) is now a reality. It is known as the universal mobile telecommunication system (UMTS)/international mobile telecommunication-2000 (IMTS-2000) (UMTS/IMT-2000). The UMTS is an initiative from the European Telecommunication Standards Institute (ETSI), and the IMT-2000 belongs to the future public land mobile telecommunication system (FPLMTS) of the International Telecommunication Union (ITU) [3]. UMTS/IMT-2000 aims to offer universal services of up to 2 Mbps in a global environment with a single subscriber number independent of service providers. UMTS/IMT-2000 will include multimedia services and must be accessible to the mass market. These objectives imply an integration between the terrestrial UMTS (T-UMTS) and the satellite UMTS (S-UMTS) components. S-UMTS will allow global coverage, enabling UMTS users to roam globally with a quality of service (QoS) comparable to that of the T-UMTS and boosting telecommunication services in developing countries. Right now the challenge is to achieve a standard S-UMTS air interface and to develop a global S-UMTS access network [4]. WATM is a technology that has been developing in cooperative work between the ATM Forum and the ETSI broadband radio access networks (BRANs). The aim of this work is to develop three subnetwork standardsHIPERLAN-type 2, HIPERACCESS, and HIPERLINKthat will be able to support ATM and most of the needs of broadband mobile systems [3]. Meanwhile, the Information Society Technologies Research Program of the European Union is developing the mobile broadband system (MBS). The MBS would provide the users of wireless systems access to the broadband services currently only available to users of the fixed integrated broadband communications network (IBCN)but with all of its mobility facilities [3]. Nevertheless, it is expected that users will eventually demand the development of new applications with broadband access and bit rates higher than 2 Mbps, including broadband WLANs, multimedia, and interactive broadcasts in a global environment based on terrestrial and satellite systems. This necessity will give rise to a fourth generation. However, the scarcity of available spectrum will pose serious obstacles to the development of the fourth generation.

6


Another important aspect of the development of wireless communications systems is fixed wireless services, or wireless local loop (WLL) services, which are used to provide local telephone services using a wireless link instead of the traditional copper wires. The main advantage of WLL is savings in installation costs, because it is not necessary to lay cables underground. This allows an easy and fast way to add new subscribers to the telephone network infrastructure at a low installation and maintenance cost. WLL is also used to replace old network segments [2]. As far as wireless terminal performance is concerned, a change in functionality is expected. Bearing in mind the applications and services that new wireless networks will offer, wireless terminals will have to evolve from their traditional formats toward new integrated solutions able to manage all the possible services available to the user. There is currently a wide variety of terminals that can be used depending on the application, including cellular phones, personal digital assistants (PDAs), and handheld personal computers (HPCs). The most popular terminals are cellular phones. Cellular phones are used primarily for voice communications, but they also incorporate basic tools for number storage. PDAs are used as electronic diaries. PDAs allow users to maintain records of appointments, personal notes, lists of contacts, and other such information. These devices do not have a keyboard for data input. Instead, they use a pen input or handwriting recognition. They have communication facilities via a serial port for data interchange with desktop computers, and some of them allow users to incorporate a modem. HPCs, which are used as portable personal computers, are also very popular devices. They include applications for word-processing presentations and facilities for communications via serial and parallel ports, modems, LAN cards, or IR ports. It can thus be concluded that there are two kinds of wireless terminals: terminals for communication, such as cellular telephones, and terminals for data processing, such as PDAs or laptop computers. The wireless communication services that are emerging today, however, are leading the way for a new generation of terminals that include both functions: communications and data processing. These new terminals will offer data processing capabilities with a cellular phone and wireless modem in a single unit. It will be crucial to consider several factors in the design and development of these terminals, including power consumption, adaptability and reconfigurability to the different possible applications supported by the terminal, security, and the user interface.

Introduction

7

There are two tendencies in the design of terminals of the new generation: ω

A classical configuration based on a central processing unit (CPU), which will manage the data-processing and communication functions but which will be subject to a bottleneck problem in the CPU;

ω

Configurations such as the wireless adaptive network device (WAND) proposed in [5]. In this type of architecture, the CPU is shut down most of the time, and the peer-to-peer data transfers are done directly without involving the CPU. Similarly, the tasks related to the data processing of audio and video streams are not done at the time-shared, general-purpose CPU.

It will be necessary to wait for the development, design, and evolution of these terminals before knowing what the basic working tool of the 21st century will be. This book will present an analysis of the standards and applications that have been developed over the last few years for wireless data communications in local environments. Cellular telephony is outside the scope of this book. This book outlines the evolution of the communication systems that will become an important part of our lives at home and in the work environment.

References [1]

Prasad, R., J. Schwarz Dasilva, and B. Arroyo-Fernández, Air Interface Access Schemes for Wireless Communications, IEEE Comm., Vol. 37, No. 9, Sept. 1999, pp. 104105.

[2]

Zeng, M., A. Annamalai, and V. K. Bhargava, Recent Advances in Cellular Wireless Communications, IEEE Comm., Vol. 37, No. 9, Sept. 1999, pp. 128138.

[3]

Prögler, M., C. Evci, and M. Umehira, Air Interface Access Schemes for Broadband Mobile Systems, IEEE Comm., Vol. 37, No. 9, Sept. 1999, pp. 106115.

[4]

Taaghol, P., et al., Satellite UMTS/IMT2000 W-CDMA Air Interfaces, IEEE Comm., Vol. 37, No. 9, Sept. 1999, pp. 116126.

[5]

Lettieri, P., and M. B. Srivastava, Advances in Wireless Terminals, IEEE Personal Communications, Vol. 6, No. 1, February 1999, pp. 619.

8


Selected Bibliography Chan, M. C., and T. Y. C. Woo, Next-Generation Wireless Data Services: Architecture and Experience, IEEE Personal Communications, Vol. 6, No. 1, February 1999, pp. 2033. Fasbender, A., et al., Any Network, Any Terminal, Anywhere, IEEE Personal Communications, Vol. 6, No. 2, April 1999, pp. 2230. Haartsen, J. C., The Bluetooth Radio System, IEEE Personal Communications, Vol. 7, No. 1, February 2000, pp. 2836. Negus, K. J., A. P. Stephens, and J. Lansford, Home RF: Wireless Networking for the Connected Home, IEEE Personal Communications, Vol. 7, No. 1, February 2000, pp. 2027. Ozugur, T., et al., Next-Generation Indoor Infrared LANs: Issues and Approaches, IEEE Personal Communications, Vol. 6, No. 6, December 1999, pp. 619. Williams, S., IrDA: Past, Present, and Future, IEEE Personal Communications, Vol. 7, No. 1, February 2000, pp. 1119.

CHAPTER

2

Contents 2.1 Introduction to the IrDAGeneral description 2.2

Physical layer (SIR)

The IrDA Standard A. Santamaría, J. R. Vento-Álvarez, J. A. Rabadán, and R. Pérez-Jiménez

2.3 Serial infrared link access protocol (IrLAP) 2.4 IRDA link management protocol (IrLMP) 2.5 IRDA transport protocol: TinyTP 2.6 LAN access extensions for link management protocol: IrLA

2.1 Introduction to the IrDA General description The IrDA is an association of over 160 companies worldwide focused on providing standards for wireless IR communications. Its aim is to ensure serial data interconnection through free-space IR channels featuring interoperability, low cost, low power consumption, half-duplex transmission, pointto-point configuration, and flexibility for easy adaptation to a wide range of appliances and devices. The IrDAs Web site may be accessed at www.irda.org. Typical areas for IrDA solutions include diagnosis and information gathering using handheld devices in automotive, medical, and industrial environments as well as telephones, PDAs, PCs, and notebook computers and printers [13] (see Figure 2.1). The IrDA has developed standards for IR communications. In September 1993, the 9

10


Application API Operative system IrDA 1.1

UART

Figure 2.1

IrComm IAS Tiny TP IrLMP IrLAP IrDA controller IrPHY

IrDA 1.1 Transceptor

IrDA protocol stack and typical application.

IrDA determined the basis for the IrDA serial IR (SIR) data link standard. In April 1994, the IrDA approved IrDA standard version 1.0, which includes the SIR link specification, the IR link access protocol (IrLAP) specification, and the IR link management protocol (IrLMP) specification. The IrDA 1.0 standard was designed to support the characteristics of ad hoc or walk-up connectivity applications as well as the unique requirements of the IR-base data link. The first IrDA communication release supported speeds from 2,400 bps to 115 Kbps and could be implemented using standard universal asynchronous receiver-transmitter (UART). In 1995, several market leaders announced or released products with IR features based on IrDA standards [4, 5]. These products included components, adapters, printers, PCs, PDAs, notebook computers, LAN access, and software applications. In November 1995, the Microsoft Corporation included support for IrDA connectivity in the Microsoft Windows 95 operating system, enabling low-cost wireless connectivity between Windows 95based PCs and peripheral devices. In October 1995, the IrDA approved version 1.1, an extension to the physical-layer standard, which provided for three new speed capabilities of 576 Kbps, 1.152 Mbps, and 4 Mbps. As these higher speeds could not be implemented on UART, it took longer to develop and incorporate technical solutions. Another important extension approved at that meeting was the IR communications protocol (IrCOMM). IrCOMM is an upper layer of the telecommunications stack targeted at emulating serial

The IrDA Standard

11

or parallel port cable protocols. With IrCOMM, the installed applications that already use a serial or parallel port to make a connection would be able to use the IR port without modification. The IrDA port is simple and easy-to-use and supports the walk-up, ad hoc connection. An IrDA 1.1 connection supports data transmission speeds from 2,400 bps to 4 Mbps. The subsystem is made up of four elements: an IrDA-enabling application, an IrDA 1.1 software protocols stack, an IR controller, and an IR transceiver. The interfaces between these elements are not defined by the IrDA standard. IrDA 1.1 controllers are available as a set of chips from several IC manufacturers or built into the wireless IR communications system inside the equipment. The IrDA is designed to be power-efficient so that it will not be a drain on the battery of portable devices such as notebook computers. The angular spread of the IR beams does not require precise alignments between the handheld device and the target device to achieve an IR link. The standard was developed to utilize low-cost components with implementation costs for manufacturing envisaged at only a few dollars per device. 2.1.1

The standard protocol stack

The layer structure over the physical level in the IrDA standard presents a fairly simple protocol structure that is open to new applications needs. Therefore, those parts near the physical level (IR link) have stable and definitive configurations. The IR link is a constant element in IrDA, so the basic control will not need many changes. This first level implementation is designed to be general and simple to make the future evolution of the upper levels easier. On the other hand, the levels nearer the applications have fewer design restrictions. They mainly describe the way of accessing the services provided by lower levels. Nevertheless, inside the different protocols there exist optional components for adapting the protocol complexity to the functions of each device. For example, a printer will have simple protocols limited to access and connectivity, while a network server will include much more complex applications. The IrDA structure can be divided as follows: ω

The physical level manages the IR transmitter and receiver and performs error detection. There are other IrDA publications related to the physical layer (PHY). The IrDA Serial Infrared Physical Layer Measurements Guidelines present the recommended tests and measurements for self-certification by developers of IrDA

12


equipment. The document includes examples of test and measuring circuits, recommended types of measurement and instrumentation, recommended calibration, and recommended test procedures. On the other hand, the IrDA Infrared Dongle Interface presents the specifications of connectors and pin assignments that allow interoperation between IR dongles and desktop systems. Readers interested in these topics can obtain information from www.irda.org. ω

IrLAP is responsible for communication between different machines. It implements station discovery processes and address assignment. It also controls the QoS negotiation procedures in established links. IrLAP supports two possible communications modes: connectionless-oriented services, which implement a broadcast transmission, and connection-oriented services, which are based on a token system.

ω

IrLMP optimizes the use of the IR link providing the possibility of several logic channels that share the same IrLAP link service. It also introduces an information service that allows different machines to see the services available in other devices.

ω

There are several optional transport protocols implementing different packet flow control algorithms. These can be selected by the user for communications through IrLMP channels.

ω

The IrLAN protocol uses services of IrLMP and transport protocols to create a LAN structure using IrDA links.

ω

Applications are any software that implement specific work or user-defined services. These applications use the services of the protocols mentioned before for their communications. Among them there are several IrDA association initiatives that are worth mentioning: IrCOMM, a protocol emulating the PC serial and parallel ports making use of TinyTP (TTP) or IrLMP services; IR transference protocol (IrTRAN-P), which is oriented to image transference through the IR channel; and IrDA plug-and-play (IrDAPNP).

Finally, the IrDA LITE protocol is also worth mentioning. The IrDA LITE protocol presents an IrDA link with less complex software and

The IrDA Standard

13

hardware. Devices with simpler communications needs should use this protocol. Next, the physical layer of the SIR specification is presented.

2.2

Physical layer (SIR) [6]

2.2.1

General description of the physical layer specification

The serial IR link is designed to support a half-duplex wireless point-topoint link between two independent nodes (such as a laptop and a printer, or two computers). The length of the link has to be between 0m and 1m, but it can work at greater lengths. The bit error ratio (BER) must be less than 10−8, and the optoelectronic technology must be low-cost. The transmitter and receiver are limited to a field of view of 30 deg (see Figure 2.2). The IrDA IR interface performs a serial transmission. The bit rate is 9.6 Kbps, but other bit rates are permissible. The pulse duration, codification, and modulation of the optical signals depend on the bit rate. The following sections detail these characteristics: Section 2.2.2 sets out the signals characteristics at bit rates from 2.4 Kbps to 115.2 Kbps. Section 2.2.3 sets out the signals characteristics at bit rates of 0.576 Mbps and 1.152 Mbps, and Section 2.2.4 sets out the signals characteristics at bit rates of 4 Mbps. Other features of IrDA systems, established in the standard, are the characteristics of the IR transmitted signal and the optical signal on the receivers side. These characteristics are detailed in Section 2.2.5.

15°30° Node 2 Optical port

01 m

Optical port Node 1

Figure 2.2

Typical connection.

14


2.2.2 Specifications for signals at bit rates from 2.4 Kbps to 115.2 Kbps

Figure 2.3 shows the block diagram of a one-end serial IrDA transmitter at bit rates up to 115.2 Kbps. The serial data to be transmitted come from a UART. Data in to the UART come from a generic source that is not the aim of the IrDA standard. The electrical signal delivered by the UART consists of groups of 10 bits, including one start bit, eight data bits (without a parity bit), and one stop bit. The encoder block transforms this signal into an electrical IrDA frame, where a 0 is encoded as an electrical pulse, and a 1 is encoded as no pulse. The pulse width is 3/16 of the bit time. The transducer module transforms the electrical IrDA frame into an optical signal, using an IR emitting diode (IRED) with a driver. Figure 2.4 shows an example of IrDA signals for bit rates up to 115.2 Kbps. During signal reception, the opposite process occurs. A photodiode receives an optical signal and transforms it into an electrical signal. This signal has the IrDA frameencoded format. After that, the signal passes through the decoder where it is transformed into the electrical 10-bit stream that can be delivered to the UART. As previously explained, the pulse width depends on the bit time of the transmitted signal. Table 2.1 presents a summary of pulse duration for data rates from 2.4 Kbps to 115.2 Kbps.

IR transducer module

Data to be transmitted UART device Received data

IrDA encoder

IRED and Driver

IrDA decoder

Photodiode and receiver

Electrical signal delivered by/to the UART Figure 2.3

Electrical IR frame

A one-end serial transmission block diagram.

IrDA optical signal

The IrDA Standard

Electrical signal delivered by/to the UART

15

Start bit

0

Stop bit

Data bits

1

0

1

0

0

1

1

0

1

Electrical IrDA frame Pulse Width = (3/16)·Bit Time

Bit Time

Figure 2.4

IrDA signals for bit rates up to 115.2 Kbps.

Table 2.1 Signaling Rate and Pulse Duration Specifications Signaling Rate

Pulse Duration (3/16)x Bit Time

002.4 Kbps

78.13 µs

009.6 Kbps

19.53 µs

019.2 Kbps

09.77 µs

038.4 Kbps

04.88 µs

057.6 Kbps

03.26 µs

115.2 Kbps

01.63 µs

2.2.3 Specifications for signals at bit rates of 0.576 Mbps and 1.152 Mbps

For data rates of 0.576 Mbps and 1.152 Mbps, the encoding method is similar to the method used for 115.2 Kbps data rates. When an IrDA link negotiates data rates of more than 115.2 Kbps, and once the connection between the two communicating nodes has been established, the system emits a SIR interaction pulse (SIP). The SIP is used to quiet down slower systems that can interfere with the link. The SIP is sent at least once every 500 ms during the connection. The SIP is an optical pulse with a duration of 8.7 µs, where the emitter is on during the first 1.6 µs; it turns off for the remaining 7.1 µs.

16


Data at 0.576 and 1.152 Mbps are sent in frames that consist of two beginning (STA) flags, one address (ADDR) field, the data, a 16-bit cyclic redundancy check (CRC), and an end (STO) flag. Figure 2.5 shows the frame format. The STA and STO are equal to 01111110 binary. The ADDR and DATA fields are defined in the IrLAP layer (see Section 2.3). The DATA field includes a CONTROL field (8 bits in length) and an INFORMATION field, which is not mandatory. The maximum length of the INFORMATION field is 2,045 bytes. The process for building the frame starts from the bytes delivered by the upper layer (IrLAP), which are listed starting from the LSB of each byte. First, the 16-bit CRC is computed for the bytes making up the address and data fields. The CRC calculation is carried out from the least significant bit to the most significant bit of each byte. The CRC (Comité Consultatif International Télégraphique et Téléphonique (CCITT)) polynomial is as follows: CRC ( x ) = x 16 + x 12 + x 5 + 1

(2.1)

Once calculated, the CRC is added at the end of the data. After that, the zero insertion occurs. A 0 is added after five consecutive 1s to distinguish the STA and STO flags from the data. Zero insertion is carried out

Bits are listed from the least significant bit to most significant bit (for each byte)

Fields

STA

STA

01111110

01111110

Bits stream sent: 01111110

Defined in IrLAP Control

ADDR

Information (optional)

CRC

STO

CRC-16

01111110

01111110 xxxxxxxxxxxxxxxx .......................... xxxxxxxx

01111110

Bits including zero insertion IrDA frame

0

Bit time

1

1

1

1

1

1

0

0

1

1

1

1

0

Pulse width = (1/4)·Bit time

Figure 2.5 Frame format for 0.576- and 1.152-Mbps signal rates. Bits are listed from the LSB to MSB (for each byte).

The IrDA Standard

17

Table 2.2 Signaling Rate and Pulse Duration Specifications Signaling Rate

Pulse Duration (1/4)x Bit Time

0.576 Mbps

434.0 ns

1.152 Mbps

217.0 ns

in the address, data, and CRC fields. After that, two STA and one STO flags are added, and the frame is ready to be transmitted. The encoding method used to transmit the frame is the same as the one used for data rates of up to 115.2 Kbps, but with a pulse duration of a quarter of the bit time. The pulse duration for these data rates is presented in Table 2.2. The presence of obstacles in the line of sight (LOS) between the emitter and the receiver or the presence of IR noise can cause an interruption in the transmission. The transmitter can also interrupt it intentionally. When a frame is interrupted, it is called an aborted frame. The receiver considers that an aborted frame has arrived when it receives seven or more consecutive 1s, or when no signal is present for seven or more bit periods. 2.2.4

Specifications for signals at bit rates of 4 Mbps

The higher data rate established in the IrDA standard, for the moment, is 4 Mbps. In this case, pulse position modulation (PPM) is used. Using PPM, bits are grouped together in sets of n bits, thus creating a data symbol. The data symbol duration, Dt , is n*Tb, Tb being the bit time. An n bits group has 2n possible combinations. Then, the data symbol duration is divided into 2n slots, called chips. Every possible combination of the n bits is codified as a pulse in one of the chips. Each pulse position corresponds to a combination of bits. The advantage of this modulation scheme is that it allows a group of bits (data symbol) with a single pulse to be sent. Pulse presence is assured independently of the bit stream, which facilitates the synchronization on the receivers side. The transmitted pulse is shorter in time, so its amplitude can be increased. Relating this to an optical signal, the power of the light pulse can be increased without damage of the IRED mean current, and greater ranges can be reached. The main disadvantage of this method is the necessity of a

18


perfect synchronization on the receivers side to decode incoming data without errors. Since 4-Mbps IrDA links work with 4PPM, bits are grouped together in twos. Thus, data symbol duration is D t = 2 × T b = 500 ns

(2.2)

Grouping two bits together, there are 22 = 4 possible bit combinations. As a result, the data symbol duration is divided into four chips. The chip duration (Ct) is: C t = D t / 4 = 125 ns

(2.3)

The correspondence between a group of bits and the position of the pulse in the data symbol is given in Table 2.3. 4PPM codification of data is done with the least significant bit (LSB) first. Bytes are represented as a stream of eight bits where the first bit is the most significant bit (MSB): b7

b6

b5

b4

b3

b2

b1

b0

4PPM is done LSB-first, so the resulting data pairs (DBPs) are 4PPMencoded into data symbols (DDs), as follows DD (of DBP b0b1)

DD (of DBP b2b3)

DD (of DBP b4b5)

DD (of DBP b6b7)

Once the bits are 4PPM-modulated, data symbols are sent starting with the data symbol of DBP b6b7: DD (of DBP b6b7)

DD (of DBP b4b5)

DD (of DBP b2b3)

DD (of DBP b0b1)

Table 2.3 4PPM Data Symbol for Each Group of Bits Group of Bits

Data Symbol

00

1000

01

0100

10

0010

11

0001

The IrDA Standard

19

Figure 2.6 gives an example of 4PPM. 4PPM encoding of bytes to be transmitted is done LSB first. Having four chips in every complete symbol, there are 16 possible combinations for filling the chips (from 0000 to 1111). However, only four of them are legal data symbols (the four combinations specified in Table 2.3). Other symbol combinations are used for special signaling. As they are not data symbols, let us call them special symbols. Data at 4 Mbps are packed into frames that include a preamble (PA) field, a STA flag, the DD field, and a STO flag, as shown in Figure 2.7. The preamble is necessary to allow clock synchronization on the receivers side. Once the preamble has been received and the receiver is properly synchronized, the STA flag allows the receiver to establish the data synchronization. Once the STA flag has been received, the receiver knows that the following pulses are data symbols that must be demodulated. The receiver stops the data recognition when a STO flag is received. Thus, the STO flag indicates the end of the frame. The DD field includes the information delivered to the PHY layer by the IrLAP layer, 4PPM-modulated (that is, only the legal data symbols previously explained can form this field). The information delivered by the IrLAP includes the destination address, control field, information, and a frame check sequence (FCS) of 32 bits (CRC32). Its maximum length is Original bytes to be sent: X27 - XD8 Original bits: 0010 0111 - 11011000 LSB-first bits stream: 11 10 01 00 - 00 01 10 11 4PPM data symbols stream: 0001 0010 0100 1000 - 1000 0100 0010 0001 Transmitted stream

1000 0100 0010 0001 - 0001 0010 0100 1000

Transmitted stream

First bit transmitted

Ct Dt=2Tb Figure 2.6

Frame format for 0.576 and 1.152 Mbps signal rates..

20


From IrLAP

ADDR

Control

Data

CRC-32

4PPM codification

PA

STA

Figure 2.7

Data symbols obtained after 4PPM codification

STO

Frame format for 4 Mbps.

negotiated during the link establishment. The CRC32 polynomial is as follows: CRC(x) = x 32 + x 26 + x 23 + x 22 + x 16 + x 12 + x 11 + x 10 + x8 + x 7 + x5 + x 4 + x2 + x2 + x + x + 1

(2.4)

Once calculated, the eight bytes forming the CRC are first inverted; then, they are 4PPM-codified using the same procedure as that used for data bytes. As a result, the inverted bytes of CRC are codified LBS-first and sent. For example, let us say that the resulting CRC-32 calculation of a given piece of data is: X36 XA7 XDE X86. The bits representation of this CRC is 00110110 10100111 11011110 10000110 The inverted bits stream is 11001001 01011000 00100001 01111001 The LSB-first representation of the previous bits stream is 10010011 00011010 10000100 10011110 Data symbols for the previous bit pairs are 0010-0100-1000-00011000-0100-0010-0010 0010-1000-0100-10000010-0100-0001-0010

The IrDA Standard

21

and the resulting DD stream transmitted is 0001-1000-0100-00100010-0010-0100-1000 1000-0100-1000-00100010-0001-0100-0010 The contents of the fields inserted by the PHY level to form the frame are now presented. The PA field consists of a set of special symbols that is formed by the repetition of the following stream 16 times:

1000 0000 1010

First chip transmitted

1000

Last chip transmitted

The STA flag consists of the transmission of the following set of special symbols:

0000 1100 0000 1100 0110 0000 0110 0000



The STO flag consists of the transmission of the following set of special symbols:

0000 1100 0000 1100 0000 0110 0000 0110



Packet transmission can be aborted by sending two or more illegal 0000 symbols or by sending any illegal symbol different from those that make up the STO field. Packets must be aborted after an STA flag transmission but before a STO flag. 2.2.5

Optical interface characteristics

The optical signal emitted in IrDA links has the following characteristics:

22

Wireless LAN Standards and Applications ω

The optical wavelength for transmission lies within the 850 900-nm band, which can be generated by the use of inexpensive IREDs.

ω

The emitted optical power per unit solid angle may range from 40 mW/Sr to 500 mW/Sr for data rates of up to 115.2 Kbps and from 100 mW/Sr to 500 mW/Sr for data rates of more than 115.2 Kbps.

ω

The transmitter half angle must be between 15 and 30 deg.

ω

The IRED switching times from on to off and from off to on depend on the data rate and must be fast enough to assure proper signal transmission.

The optical characteristics for the received signals are described as follows. ω

A properly functioning IrDA receiver must be able to detect and use a signal with irradiance levels of between 4 and 500 mW/cm2. The irradiance is defined as the received power per unit area (in milliwatts per square centimeter).

ω

The maximum optical power delivered to the detector will be reached by a source operating at the maximum optical power with a minimum link length. These conditions must not cause receiver overdrive distortion and possible related link errors.

ω

The minimum optical power at the receiver will occur while operating at the minimum power and with the maximum link length. Under these conditions the receiver must meet the BER specifications.

For more information on these characteristics see [6].

2.3

Serial infrared link access protocol (IrLAP)

The IrLAP specification [7] describes the functions, features, protocol, and services for interconnection between computers at the data link layer [open system interconnection (OSI) layer 2]. It is based on the well-

The IrDA Standard

23

known asynchronous high-level data link control (HDLC) and syncronous data link control (SDLC) [812] half-duplex protocols with some major modifications, which can be summarized as follows: ω

Addressing is extended to take into account the ad hoc nature of the medium. It also means that dynamic address conflict resolution and station discovery/identification procedures have to be introduced.

ω

Connection setup is extended to include a negotiation framework that stations use to establish the best connection characteristics that both connecting parties can support. It also means that any station can contend to become a primary station.

ω

Recovery mechanisms are extended to account for the ad hoc nature of the medium.

ω

Medium access rules are extended to resolve contention between stations competing for control of the medium and to prevent hidden-node transmissions.

IrLAP provides two general types of services to the upper layer: connectionless services and connection-oriented services. They are specified in terms of service primitives (describing the service provided but not the means by which the service is provided) and parameters. These primitives can be classified into four generic types (Figure 2.8): 1. Request: Passed from the upper layer to invoke a service. 2. Indication: Passed from IrLAP to the upper layer to indicate an event or to notify the upper layer of an IrLAP initiated action. 3. Response: Passed from the upper layer to acknowledge some procedure invoked by an indication primitive. 4. Confirm: Passed from IrLAP to the upper layer to convey the results of the previous service request. IrLAP uses these primitives to communicate with the upper layer to manage the communications processes in the link between devices.

24


Upper layer

Request

Upper layer

Response

Confirm

IrLAP layer

Indication

IrLAP layer

Frame(s) Frame(s) Figure 2.8

2.3.1

IrLAP services and primitives structure.

Services provided by IrLAP: Connectionless services

This section discusses three main services: discover adjacent stations, address conflict resolution, and data broadcasting. Discovery service helps users to determine what, if any, devices are within communication range and available for connections. A list of the available devices is returned to the emitter, and information on the device that issued the request is given to the receiver. Address conflict services are used to resolve device address conflicts. If a discovery operation yields a discovery log that contains entries for more than one device with the same device address, this service primitive may be invoked to cause the IrLAP layers of the conflicting devices to select new nonconflicting device addresses. Unit data services provide a way of transmitting data outside of a connection. Data are sent broadcast and cannot be directed to a specific device address. 2.3.2

Services provided by IrLAP: Connection-oriented services

There are six services: ω

Connect is used to request that an IrLAP connection be established to a station with a device with a specified address and a given QoS. This connection can be in a sniff mode (special low power connect procedure) or not.

ω

Sniffing is used to initiate or cancel the sniff mode.

The IrDA Standard

25

ω

Data services can either be sent as sequenced or unsequenced data. No confirmation is returned to the sender (even for sequenced data).

ω

Status is used by IrLAP to inform the upper layer that the quality of the link is suspect. Either the link is experiencing high levels of noise or all connection activity has ceased. If the link quality does not improve, then IrLAP disconnects automatically.

ω

Reset causes all unacknowledged data units to be discarded. All counters and timers are reset. A reset only occurs if both ends of the connection agree to it.

ω

Disconnection request terminates the logical connection, and all outstanding data units are discarded. No confirm primitive is needed since the disconnection is always successful.

2.3.3

Configurations and operating characteristics

Data link states A data link channel can be in one of two basic states: connection (when two or more nodes have an established connection and are exchanging control and/or information frames) and contention (when a connection is disconnected or when no connection ever existed). The IrLAP protocol treats the IrDA SIR medium as an unbalanced data link due to its half-duplex nature (without collision detection). It involves two or more participating data stations. One station (primary) assumes responsibility for the organization of data flow and for unrecoverable data link error conditions. The frames transmitted by the primary station are known as command frames. The other stations on the data link are known as secondary stations, and their transmitted frames are response frames. All transmissions through the data link go to or from the primary station, which is unique. Not all stations must have primary capability (but usually do), but those that do not can only communicate with stations that have primary capability. Modes IrLAP data stations can be in either of two modes: normal response mode (NRM) or normal disconnect mode (NDM) corresponding to the connection and the contention state, respectively. After entering NRM, each station knows which role it is to play: primary or secondary station. A secondary station will initiate transmission only as a result of receiving explicit permission to do so from the primary station. After receiving permission, the secondary one will initiate a response

26


transmission consisting of one or more frames. The secondary station will also indicate explicitly the last frame of the response transmission. Following an indication of the last frame, the secondary station will stop transmitting until explicit permission is again received from the primary station. Communications in NDM are contention-based. As a result, stations that wish to transmit while in NDM must use caution and follow the NDM media access rule strictly. 2.3.4

IrLAP frame structure

Each IrLAP frame has two or three fields (see Figure 2.9): Address (A) identifies a secondary station connection address; control (C) specifies the function of the particular frame. In addition, there is an optional information (I) field that contains the information data. Each of these fields contains either eight bits or a multiple of eight bits. Together, the A, C, and I fields are referred to as the payload data. Wrapping layer Each IrLAP frame is preceded and succeeded by fields that constitute the wrapping layer, which implements a PHY scheme that serves to transmit the payload data reliably. The wrapping layer fields serve to mark the beginning and end of the frame and to check for the reliable transmission of data. The format of the wrapper fields will vary according to the particular PHY scheme used, but every frame wrapper will include at least three components: a STA flag, a FCS field, and a STO flag. 2.3.5

IrLAP frame types

The type of frame or function is determined from the content of its C field. There are three general C field formats: unnumbered (U), information (I), and supervisory (S) format. Unnumbered (U) format U frames are used for data link management (see Figure 2.10). This function includes establishing and disconnecting the

Address

Control

Information

8 bits

8 bits

8 * M bits

First byte of payload data delivered to/received from the physical layer Figure 2.9

IrLAP frame structure.

The IrDA Standard

Figure 2.10

27

7

6

5

3

2

1

0

X

X

X P/F X

X

1

1

4

IrLAP U frame structure.

data link and reporting procedural errors or transferring data (when the location of the data in a sequence of frames is not to be checked). The values of the five bits marked as X can be coded to determine the nature of the U frame. This will be described in more detail in the poll/final (P/F) bit section. S frames assist in the transfer of information, although they do not carry information themselves. They are used to acknowledge received frames, to convey ready or busy conditions, and to report frame-sequencing errors. Supervisory (S) format

Information (I) transfer format I frames transfer information. The I field contains data that is moved from place to place in the system. This I field is unrestricted in content.

While the frame is the vehicle for data transmission, the higher grouping level is called frame sequence. To avoid frame repetition, each station maintains two state variables: Vs and Vr. Vs denotes the sequence number of the next sequenced frame to be transmitted. Vr denotes the sequence number of the next sequenced frame expected to be received. The C field of the I-sequenced frames also contains send and receive counts (Ns and Nr, respectively). IrLAP uses Ns count (which indicates the number of the I frame within the sequence of I frames transmitted) to ensure that these frames are received in their proper order. The transmitted frame uses Nr count to confirm that received I frames are accepted. Frame sequencing

The P/F bit occupies bit 4 of the C field in U, S, and I format frames. The P/F bit is used to control the two-way alternative access to the link when in a connection (NRM). When sent from a primary station, it is the poll (P) bit. The primary station uses this to solicit a response or sequence of responses from the secondary station. When sent from a secondary station, it is the final (F) bit, which is used by the secondary station to indicate the final frame transmitted as the result of the previous poll command. This bit can be seen as a mechanism for giving The poll/final (P/F) bit

28


transmissions permission on the link when in NRM. The secondary station is not allowed to transmit until it receives a command frame with the P bit set to 1. The secondary station may then send multiple frames to the primary station. The secondary sets the F bit to 1 when sending the last frame of its response transmission. This gives transmit permission back to the primary station. At that point, the secondary station no longer has permission to transmit on the link.

2.4

IRDA link management protocol (IrLMP)

IrLMP [13] is the upper level of IrLAP. It is based on the IrLAP layer and provides services to the protocol and applications levels. It has two main tasks associated with each of its components: ω

First, IrLMP maintains a database that contains all the services available to other machines. Other devices can test this component, the link management information access service (LM-IAS), to determine which service each machine provides.

ω

Another IrLMP main component is the link management multiplexor (LM-MUX). It is responsible for the implementation of several links using the only channel supplied by IrLAP. These links can be between machines (using the IrLAP channel) or between two applications in the same machine (using a connection managed by LM-MUX).

LM-IAS uses the LM-MUX services and can also be understood as an application over this level. Figure 2.11 shows the general IrLMP structure and its situation inside the IrDA standard. IrLMP (together with IrLAP and physical level protocols) are the only mandatory levels in the IrDA standard. They are all responsible for the establishment of a reliable link. Other levels are optional and can be personalized according to the application requirements. 2.4.1

Link management multiplexor (LM-MUX)

LM-MUX uses the single channel provided by IrLAP and implements several logic channels with each one independent from the others. These LM-MUX channels are used by upper-level applications in the same

The IrDA Standard

29

Applications

IrLMP

LM-IAS

Transport protocols TinyTP ...

IrCOMM IrPNP ...

LM-MUX IrLAP Physic Figure 2.11

General IrLMP structure within the IrDA standard.

device or in different terminals and with one of two possible configurations: multiplexed or exclusive mode. In multiplexed mode there are several available connections that are routed through the IrLAP link. LM-MUX multiplexes the packets from the applications and gives them to the lower-level service. It also takes the received packets and sends them to the corresponding upper-level client. LM-MUX has no flow control of the packets between applications, presenting a highly simplified structure and allowing the use of any flow control protocol. Exclusive mode permits only one connection, using all the IrLAP link capabilities. In this case it is possible to use the flow control implemented in IrLAP as a base for the flow control of the application. LM-MUX external interface The IrDA standard does not make restrictions on the protocol implementations but details how the protocols must communicate between them. LM-MUX provides its services by means of LM-MUX services access points (LSAPs). There will be an LSAP for each application, presenting a different LSAP connection endpoint for each connection made by the application. Furthermore, in the IrLAP layer a unique IrLAP services access point (ISAP) will exist, containing several ISAP connection endpoints, one for each link with different machines. Consequently, LSAP connection endpoints are connected to their peers in the other machines through the ISAP connection endpoints. It is also possible to connect LSAP connection endpoints on the same device to

30


communicate different applications running on it. Figure 2.12 shows an example of these connection schemes. Each LSAP has an identifier called LSAP selector (LSAP-SEL), whose value is different for each LSAP in the machine. Frames generated in a connection should contain the LSAP-SEL of the origin and destination LSAP, so as to allow LM-MUX to route the frames correctly using the LSAP-SEL as address. There are two special LSAPs with only one LSAP connection endpoint each. A connectionless LSAP takes the connectionless messages of the LSAPs and indicates this event to all the LSAPs connected to them. The other particular LSAP is XID_Discovery LSAP, which manages the discovery and sniffing process. LM-MUX internal organization The services provided by LM-MUX, and the IrLAP services that they require, are connected through several finite state machines (FSMs). Each FSM manages LSAP and received packets and controls the general LM-MUX processes. Figure 2.13 shows the internal LM-MUX structure and its different components: LSAP control FSM, receptor/demultiplexor, and station control. LSAP control FSM carries out both LSAP connections and disconnections. This control is carried out by means of a channel request to the station control and a connection request to the remote LSAP control FSM. LSAP c. e. LM-MUX Service boundary

LSAP

LrLAP Service boundary

ISAP

Physic boundary

ISAP c. e.

IrLAP connection LSAP connection

IrLAP Layer

IrLAP Layer

LM-MUX Layer

LM-MUX Layer LM-MUX Clients

Figure 2.12

LM-MUX Clients

Example of LM-MUX external connections.

The IrDA Standard

31

LM-MUX Service boundary

Connectionless XID_DISCOVERY LSAP LSAP

LSAP-conn. control FSM

Station control RX/DMUX

IrLAP Service boundary

Figure 2.13

Internal LM-MUX structure.

The receptor/multiplexor delivers incoming packets from the IrLAP link and starts new LSAPs when connection requests are received. It uses a frame-addressing format similar to: <SLSAP-SEL> where DLSAP-SEL and SLSAP-SEL correspond to the destination and source LSAP-SELs. Finally, station control is responsible for the link control and the coordination of the LSAP connection FSMs. It is made up of several elements. The station control FSM implements discovery and sniffing processes, address resolution, IrLAP connections and disconnections, assignment of LSAP connections to IrLAP connections, and transitions between exclusive and multiplexed mode. Station control also maintains an IrLAP connection control FSM for each IrLAP connection endpoint present at any moment in the ISAP, with an associated LSAP connection table containing all the LSAP connections. This FSM is also responsible for making IrLAP connections when there is a request from an LSAP associated with it, and for finishing them when there are no active LSAP connections using the IrLAP connection. Finally, the IrLAP connection table contains information on active IrLAP connections. Figure 2.14 shows the station control scheme.

32


LM-MUX Service boundary

LSAP-conn. control FSM

Connectionless XID_DISCOVERY LSAP LSAP

IrLAP LSAP-conn. control conn. FSM table

IrLAP conn. table

Station control FSM

IrLAP LSAP-conn. control conn. FSM table RX/DMUX IrLAP Service boundary Figure 2.14

Station control scheme.

LM-MUX services This item describes the services that transport protocols and applications can obtain from LM-MUX. The LSAP connection endpoints situated in the LM-MUX boundary offer the following capabilities: ω

Discovery sniffing: Based on the same type of services provided by the IrLAP level;

ω

Connection/disconnection (or breakdowns): Links the source and destination LSAPs, using, if both LSAPs are in different machines, the IrLAP link connections;

ω

Access mode: The capability to change between multiplexed and exclusive mode;

ω

Status: Allows information to be obtained on the connections (e.g., mode and QoS parameters);

ω

Data transfer: The primary expected feature. It is important to note that flow control is not implemented by these services but has to be managed by the applications. Only in the exclusive mode is part of flow control based on IrLAP flow control.

The IrDA Standard

33

LM-MUX frames are transmitted in the IrLAP frames information field. They are called LM-MUX protocol data units (LM-PDUs), and Figure 2.15 shows their configuration. In Figure 2.15, C denotes a control bit (indicating whether the frame is command -C = 1- or data -C = 0-), and r is a reserved bit for future applications; it must be set to zero. The information field has different structures depending on whether it is a command or a data frame. The specific command frames format is given in Figure 2.16, where the value of A determines whether it is a command request (0) or response (1). Frame format

2.4.2

Information access service (IAS)

Several devices connected in an IrDA framework may have different local applications available to applications running in other machines. It is possible to communicate these applications through the IR link by means of LM-MUX, but for these connections to take place it is necessary to know where each application is located and to identify the required parameters to characterize them. Information access service (IAS) provides the tools to look for available services in other machines. It also maintains an information base of the services offered in the local machine being consulted by other devices. All these processes are based on the information access protocol (IAP), a simple protocol that carries out communications between two IAS entities. Figure 2.17 shows the IAS structure; it is made up of an information base and two FSMs. They are the core of the IAP and consult other machines (client IAP) and reply to other devices requests (server IAP). To simplify information access, the server IAP is always located at the same address as the LSAP numbered zero.

76543210 76543210 C DLSAP-SEL r SLSAP-SEL Information Figure 2.15

LM-MUX frame structure.

76543210 A OPCODE Parameters Figure 2.16

LM-MUX command frame structure.

34


Local data updates IAS Interface

Client IAP FSM

Information base

IAP

Server IAP FSM

LM-MUX Service Interface Figure 2.17

IAS structure.

Object 0 Class device DeviceName IrDASupport Figure 2.18

Object 1 Class service 1

MyDevice

Attribute 1

5

Binary data

Attribute 2

String

Attribute 3

Binary

IAS class example.

IAS information model Data in the information base is structured in classes. A class is made up of objects characterized by an identification number and a number of attributes (name and type value). Figure 2.18 gives an example of this structure. The object numbered zero in the device class is always present and contains information on the local machine. IAP provides primitives to obtain information from the database (objects of a determined class or attributes values, for example). In any case, the external devices cannot modify data in the information base; only applications in the local machine are allowed to change and insert information into the local database. IAS frames format IAS frames are transmitted inside the information field of the LM-PDU frames. Figure 2.19 shows their structure. Within this structure, the control (CTL) field is made up of three elements (see Figure 2.20).

The IrDA Standard

35

CTL Figure 2.19

IAS frame structure.

7

6

LST A S K Figure 2.20

DATA

8 O P. C O DE

IAS frame C field structure.

An LST indicates whether the frame is the last one of a message (LST = 1) or not. ASK is used in the flow control and is the received-packet acknowledgment bit.

2.5

IRDA transport protocol: TinyTP

As we have seen before, LM-MUX does not provide flow control for the different LSAP connections. Only in exclusive mode is there some kind of control based on the IrLAP flow control. Therefore, a flow control is necessary in multiplexed mode, which manages each LSAP packet flow. TinyTP [14] is an optional transport protocol available for use by client applications as flow control in IrDA links. It provides a flow control for each LSAP connection and implements a segmentation and reassembly (SAR) mechanism. SAR ensures arbitrary length stream transmission. TinyTP uses LM-MUX services through an LSAP connection endpoint, whose LSAP-SEL number also identifies the TinyTP services access point (TTPSAP). Applications can use TinyTP services or implement their own flow control algorithms working directly with the LM-MUX services. It is also possible to develop different transport protocols and use them instead of TinyTP. The services provided to TinyTP clients can be summarized as follows: ω

Connection and disconnection: Between two LSAPs;

ω

Data transmission: Information interchange between two TTPSAPs;

ω

Local flow control: Stopping or resuming the packet flow.

36


2.5.1

TinyTP frames format

The TinyTP protocol data units (TTP-PDUs) are transmitted in the information field of the LM-PDUs and have the format shown in Figure 2.21, where M corresponds to the bit most used with SAR service. If this service is disabled (M = 0), all the messages must fit into a TTP-PDU. When it is activated, messages are truncated and transmitted in successive TTPPDUs and reassembled in the receptor machine SAR. DeltaCredit is used in the flow control algorithm, indicating how many TTP-PDUs can be sent from the receiver of the packet to the sending machine. In the TTP-PDUs used in the connection process, a new field with different parameters referring to that connection can appear. At its initial value, DeltaCredit is called InitialCredit. The two possibilities are shown in Figure 2.22, where P specifies whether there is (P = 1) or is not (P = 0) a parameter field. 2.5.2

Flow control

The credit system used by the TinyTP data control procedure is based on three variables that are updated from values received in InitialCredit and DeltaCredit: ω

SendCredit corresponds to the number of packets that can be sent by the local TTPSAP. This value is decreased when a TTP-PDU is delivered. When SendCredit equals zero, no packets can be transmitted.

1 bit

7 bits

M DeltaCredit Figure 2.21

TTP-PDU structure.

1 bit P

Figure 2.22

Data...

7 bits Data...

InitCredit

1 bit

7 bits

P

InitCredit

Parameters Data...

TTP-PDU connection schemes.

The IrDA Standard

37

ω

RemoteCredit is the number of packets that can be received from another remote TinyTP user. It corresponds to the SendCredit variable in the peer entity.

ω

AvailableCredit is the number of packets that could be received but have not been advanced to the other machine. As the DeltaCredit field is seven bits long, available credits can be advanced only in blocks of up to 127.

Therefore, when a user sends a packet to its peer with information and a value of available credits in the DeltaCredit field (or InitialCredit in the connection packets), SendCredit is decreased by one unit, and AvailableCredit is decreased in the DeltaCredit field value. The other user receives the packet, decreases RemoteCredit, and starts the same process again. As a result, the reception buffer can be implemented in a circular structure (see Figure 2.23). When a packet is received, it is introduced into the buffer, thereby reducing the RemoteCredit zone. Once a packet is processed, the position that it occupied is added to the AvailCredit zone. When the machine sends a packet, the RemoteCredit value is incremented in the AvailCredit zone by the value sent in the DeltaCredit field. AvailCredits

Read data

Not used RemoteCredits Figure 2.23

Flow control scheme.

Received packet queue

38


2.6 LAN access extensions for link management protocol: IrLAN The creation of the IrDA protocols and their broad industrial support has led to IrDA-compliant IR ports becoming common in laptop computers. With the IrDA approval of the higher data rates of 1.15 and 4 Mbps, the IR link is becoming fast enough to support a network interface. The main features of the IrLAN protocol [15] are described as follows. ω

It enables a computer with an IrDA adapter to connect to a LAN through an access point device that acts as the network adapter for the computer.

ω

It enables two computers with IrDA adapters to communicate as though they were attached through a LAN.

ω

It enables a computer with an IrDA-compliant adapter to be attached to a LAN through a second computer that is already attached to the LAN (the second computer must also have an IrDAcompliant adapter).

2.6.1

IrLAN general description

The IrLAN protocol is a passive, sided protocol that defines a twochannel interface between a protocol client and a protocol server. Each station has both an IrLAN client and a provider. The client begins setting up the connection by reading an objects information in the providers IAS. The object specifies an IrLMP LSAP for the control channel. The client connects to the control channel and uses the control channel to negotiate the characteristics of a data channel. Once the data channel has been negotiated, it is opened and then configured. All configurations are handled through the control channel. The data channel is used solely for the transmission and reception of packets formatted for the LAN. The IrLAN protocol defines a graceful close, but it is seldom used because it would require user intervention to initiate the disconnect procedure. Typically, the connection will close down ungracefully through an IrLAP connection time-out. Both the control and data channels use the TTP protocol for the SAR of packets and for flow control.

The IrDA Standard

2.6.2

39

Access methods

The IrLAN protocol supports three different operating modes that represent the possible configurations between IR devices and between IR devices and a connected network (see Table 2.4). An access point device is hardware supporting both a LAN network interface controller (NIC) and an IR transceiver. The access point device implements a network adapter for the client using IR as the bus for accessing the adapter. Figure 2.24 shows the access point mode of operation. Access point mode

Table 2.4 Operating Modes for IrLAN Model

Description

Access point

An IR device provides access to a LAN through the device

Peer-to-peer

Two or more computers with IR support can communicate as if they were connected through a network

Hosted

Two or more computers can communicate both with a host computer and each other as if they were all connected through a network. In addition, a physical network connected to the host is accessible to all of the computers

LAN Bridge AP

IrLAN Provider IrLMP/TinyTP

Client OS

Client OS

IrLAN Client

IrLAN Client

IrLMP/TinyTP

IrLMP/TinyTP

Client 1

Figure 2.24

IrLAN access point mode of operation.

Network NIC

Client 2

40


Peer-to-peer mode The IrLAN protocol peer-to-peer mode permits nodes to run network operating systems that are peer-to-peer capable of creating ad hoc networks. Figure 2.25 shows the peer-to-peer mode. There is no physical connection to a wired LAN. In peer-to-peer mode, each peer must provide a server control LSAP in addition to its client control LSAP and data LSAP. Each client control LSAP connects to its peers server control LSAP. This allows each node to establish and control its peers data LSAP using the command set. The IrLAN control protocol also arbitrates as to which peer initiates the data channel connection. Client OS IrLAN Client IrLAN Provider IrLMP/TinyTP

IrLMP/TinyTP IrLAN Provider IrLAN Client Client OS

Figure 2.25

IrLAN peer-to-peer mode of operation.

Client OS IrLAN Provider IrLMP/TinyTP Bridge Host

IrLMP/TinyTP IrLAN Client Client OS Station

Figure 2.26

IrLAN hosted mode of operation.

NIC

Network

The IrDA Standard

41

Hosted mode In hosted mode, the provider has a wired network connection but has multiple nodes attempting to communicate through the wired connection. Figure 2.26 shows the hosted mode. Unlike access point mode, both the host machine and the client(s) share the same NIC address in host mode. To make host mode work, the host must run special bridging and routing software that will handle the proper routing of packets. The algorithms used in this mode are highly protocol-dependent. 2.6.3

Frames size and format

TinyTP allows for the fragmentation and reassembly of packets, which may span several IrLMP frames. During the setting up of the TinyTP connection, a maximum assembled frame size is negotiated between the two sides. The IrLAN protocol currently defines support for access to the 802.3 (Ethernet) and 802.5 (token ring) LANs and is being modified to support additional media types. For Ethernet, the assembled TinyTP frame size is 1,518 bytes, while for token ring it can be up to 65,535 bytes. Data frame format The IrLAN protocol defines the commands used in the control channel as well as the format of data on the data channel. These formats are defined supposing that SAR and flow control are handled by the TinyTP interface. For 802.3 (Ethernet), the format is the same as would be transmitted at the software level for an 802.3 packet (see Figure 2.27). For 802.5 (token ring), the IrLAN data channel packet format is shown in Figure 2.28.

Destination Address [6] Figure 2.27

Access Control [1]

Source Address [6]

Length or Frame Type [2]

Information [0..1500]

IrLAN data channel packet for 802.3.

Frame Destination Source Routing Routing Information Control Address Address Control Information [1] [6] [0..2] [0.. 16] [6]

Figure 2.28

IrLAN data channel packet for 802.5.

42


Once the data channel is established, it is treated as the send and receive path for all frames on the emulated LAN media. All packets sent from a node are transmitted on this channel, and all packets being received will come from this channel. Control channel frames format The control channel is used to carry out data channel connection and configuration. It uses TinyTP as a flow control and SAR protocol. The client and provider must both support a minimum 1,024-byte assembled frame size in the control channel. If a client has to send a command that exceeds 1,024 bytes, which is highly unlikely, it must send a sequence of smaller commands of the same type for the same purpose. The control channel uses a command/response protocol. Only client-initiated command/response pairs are now defined. During a session, the client issues a sequence of request packets, each of which is immediately followed by a response from the provider. Unrequested responses are nottodaysupported by IrLAN. Table 2.5 summarizes the command

Table 2.5 IrLAN Command Description Command Code

Description

Get Provider Information

Used by the client to determine the media type/data frame formats supported by the provider and the IrLAN modes supported by the provider (access point, peer-to-peer, and/or hosted)

Get Media Characteristics

Used by the client to get detailed information on the media types supported by the provider

Open Data Channel

Used by the client to get an IrLMP LSAP number on which it should establish a TinyTP connection to the provider for the data channel

Close Data Channel

When the provider receives this command, it will stop sending packets to the data channel and stop sending received packets on the LAN; it is still up to the client to close the TinyTP connection

Reconnect Data Channel

Used by the client to reconnect a data channel; if the reconnection is successful (the provider returns a status code of zero), the state of the data channel is the same as when the channel was disconnected

Filter Configuration

Used by the client to control the filtering of packets from the provider to the client; this command also allows the client to check the filter configuration on the provider; packet filtering allows a network client to set up a description of the type of packets that the client wants to receive from the network

The IrDA Standard

43

packets available to the IrLAN client and the response packets returned by the provider.

References [1] Ingham, B., and R. Helms, Infrareds Role in Wireless Communication Expand With IrDA, Electronic Design, July 1997, pp. 6567. [2] Kahn, J. M., and J. Barry, Wireless Infrared Communication, Proceedings of the IEEE, Vol. 85, No. 2, February 1997, pp. 265298. [3] Goldber, L., Infrared Data Transmission. The Missing Link?, Electronic Design, April 1995, pp. 7375. [4] Argerstein, J., Schairer, IrDA-Compatible Data Transmission, Design Guide Temic Semiconductor, April 1996. [5] Hewlett Packard, IrDA Data Link Design Guide, Hewlett Packard Co., 1995. [6] Tajnai, J., et al., SIR-Physical Layer Specification, December 1997. [7] Serial Infrared Link Access Protocol, Version 1.1, June 1996. [8] ISO 4335 High Level Data Link Control (HDLC) ProceduresElements of Procedures 1991-09-15. [9] ISO 8885 High Level Data Link Control (HDLC) ProceduresGeneral Purpose XID Frame Information Field Content and Format, 1991-06-01. [10] ISO 3309 High Level Data Link Control (HDLC) ProceduresFrame Structure, 1991-06-01. [11] ISO 3309 Amendment 2 High Level Data Link Control (HDLC) ProceduresFrame Structure, 1991-06-01. [12] ISO 8886 Information technologyTelecommunications and information exchange between systemsData link service definition for Open Systems Interconnection, 1992-06-15. [13] IrDA._Infrared Data Association Link Management Protocol (IrLMP), www.irda.org. [14] Infrared Data Association TinyTP: A Flow-Control Mechanism for use with IrLMP, available from the IrDA, www.irda.org. [15] Infrared Data Association LAN Access Extensions for Link Management Protocol IrLAN, www.irda.org.

.

CHAPTER

3

Contents 3.1 Introduction to IEEE 802.11: General description 3.2 Medium access control (MAC) for the IEEE 802.11 wireless LANs (WLANs) 3.3 Physical layer for IEEE 802.11 wireless LANs: Radio systems 3.4 Physical layer for IEEE 802.11 wireless LANS: Infrared systems 3.5 Conclusions and applications

The IEEE 802.11 Standard R. Pérez-Jiménez, J. M. Riera, and F. J. López-Hernández

3.1 Introduction to IEEE 802.11: General description If there were a queen of LANs, it would be the Ethernet. The Ethernet is the common, and commercial, name for the ISO 8802-3, the name of the standard assuming the IEEE 802.3 recommendation. There is an enormous family of LAN standards. Almost all of them deal with just the two lower layers of the OSI architecture, the data link layer (DLL) and the PHY. In fact, some sublayers are defined to simplify the implementation of conformant equipment. In this way, the DLL is divided into the logical link control (LLC) and the medium access control (MAC), while the PHY includes the PHY convergence protocol (PLCP) and the PHY medium dependent (PMD). Other management layers have been included to optimize the layer or sublayer coordination. 45

46


The standard describes the functionality and relationships between the layers and sublayers but fails to specify how they are to be made. This is an important point, because the rules set out by the standards only deal with the behavior of the equipment, allowing manufacturers to implement it as they wish. Together with the definition of any standard, the conformance tests are included. This is the touchstone to assure interoperability between devices from different manufacturers. One of the younger members of the IEEE 802 family is the ISO 802.11 WLAN standard. It only describes the specification of the MAC and PHY layers, so wireless devices can use the same LLC developed for other IEEE 802compliant systems. Figure 3.1 describes this structure. The main goal of the 802.11 standard is to achieve full functionality for the upper layers without worrying about the quite significant differences between a network based on a reliable cable and another using the air. This benefit is possible because of the careful design of the MAC and PHY. All the issues on security, link losses, node authentication, fading, and large differences in power have been analyzed and incorporated as MAC duties or PHY services. An 802.11 network can be as simple as two stations interchanging information, and as complex as several buildings with network services in several rooms and with a connection to a backbone-cabled network. Between these two extremes there are many possibilities. The weak spot in the 802.11 standard is its complexity. Many features of this standard, needed to establish reliable data communication between several nodes, make the system so complex that the cost of the system is much higher than that of cabled networks. It is important to remember that quality is not inexpensive. Moreover, other standards,

MAC MAC-PHY Interface PHY Convergence Layer Medium Dependent PHY

Figure 3.1

Basic layer structure for the IEEE 802.11 standard.


47

such as IrDA (infrared) or Bluetooth (radio frequency), offer low-cost applications. The MAC sublayer of a LAN is responsible for correct frame transmission between stations. The IEEE 802.11 standard defines a single MAC protocol for use with all of the PHYs. This supposes that the special characteristics of both RF [1] and IR channels [25] are kept in mind. The use of a single MAC protocol better enables chip vendors to achieve highvolume production that will help keep the costs for these systems low. There was considerable debate and compromise before the adoption of the current 802.11 MAC protocol. The MAC protocol defined in the 802.11 is quite complex. The protocol has a few options, as well as several features that can be turned on and off, and combines most of the functionality that was contained in the dozen or so MAC proposals considered by the committee. The important characteristic of the 802.11 MAC protocol, which is likely to remain unchanged in the final standard, is its ability to support the following. ω

The access point (AP)oriented and ad hoc networking topologies;

ω

Both asynchronous and time-critical traffic (called time-bounded services in the 802.11);

ω

Power management.

The structure of an IEEE 802.11 network is described in the MAC section, because the functionality and duty of every station is controlled by this sublayer.

3.2 Medium access control (MAC) for the IEEE 802.11 wireless LANs (WLANs) This section briefly describes the MAC sublayer of the IEEE 802.11 WLAN standard, summarizing some of the general considerations of the MAC design and discussing the features usually found in a WLAN MAC protocol. Then, the standardized access methods, the distributed coordination function (DCF), with or without handshaking, and the point coordination function (PCF), are described. Finally, a comparison between IEEE 802.11 MAC and HIPERLAN MAC is presented.

48


3.2.1

Expected features of a WLAN MAC protocol

The 802.11 study group set out some requirements for an appropriate MAC protocol [69]. These requirements, which are summarized as follows, may generally be considered as the features expected in any WLAN, not just an IEEE 802.11 LAN. 1. Throughput: Since the spectrum is a scarce resource, throughput is definitely one of the most critical considerations in the design of a MAC protocol. The capacity of WLANs should ideally approach that of their wired counterparts. However, due to physical limitations and limited available bandwidth, WLANs are currently targeted to operate at data rates of 120 Mbps. The most extended random access protocols belong to the ALOHA family, including carrier sense multiple access (CSMA). The ALOHA family suffers from stability problems. That is, the peak throughput is accompanied by a tremendous delay. With Ethernet, and its 10 Mbps of physical transmission and over 80 percent of throughput for CSMA/CD, it is possible to deliver over 8 Mbps of theoretical performance, but, in practice, measurements show that only 33.5 Mbps performance is achieved. We should consider not only theoretical throughput but operating throughput (which, practically, is more important). One way of increasing the throughput is by using spread spectrum techniques, which support multiple transmissions simultaneously [1013]. Another important consideration about throughput is the impact of unauthorized network access. Neither the MAC nor network management functions can identify any unauthorized access before receiving its transmission, such access inevitably has an impact on network throughput and delay. A successful MAC and network security scheme should reject such unauthorized access and minimize its impact. 2. Delay: Delay characteristics are important in every application, but especially in WLANs, since they should serve not only the mandatory asynchronous data service, but also time-bounded multimedia applications such as voice and video. Delay can also cause problems for all data services where the preservation of the sequence of packets is extremely important.


49

3. Transparency to different PHY layers: One of the special requirements for a WLAN MAC is transparency to different physical transmission layers. For IEEE 802.11 LANs, physical transmission layers include direct sequence spread spectrum (DS-SS), frequencyhopped spread spectrum (FH-SS), and diffuse IR. These physical transmission layers are different not only in system design but also in propagation characteristics. However, one MAC must handle all of them. One way of achieving this goal is to have a physical dependent layer, a physical convergence layer, and an appropriate MAC-PHY interface in each station. This architecture is shown in Figure 3.1. Based on architecture currently being adopted by the IEEE 802.11 committee, a single MAC can exchange data transparently with different PHYs via the MAC-PHY interface. Directly related to this item is the limitation of the complexity of the PHY (medium dependent layer, PHY convergence layer, and MAC-PHY interface) to a minimum. The design of a WLAN is an integrated problem, from the PHY up to the network management layer. A MAC design that creates difficulty in other parts/layers of the system is undesirable. 4. Fairness of access: The fading characteristics of indoor channels may cause unequal received power at the base station even when power control is enforced. Such a situation may result in unfair access to the network. That is, one mobile node may receive much less power at the base station than another mobile node. When the MAC protocol is operating in the contention mode (necessary for initial registration and often used for uplink traffic), the disadvantaged mobile node may not have a chance to access the channel for a while. A MAC protocol should be able to resolve this situation since it is possible that capture can take place with as small as a 69 dB power difference while the dynamic range of fading can be as large as several dBs. 5. Battery power consumption: Typically, the 110V (or 220V) electrical supply provided in a building powers device connected to a wired network. Wireless devices, however, are meant to be portable and/or mobile and are typically battery-powered. Therefore, devices must be designed to be very energy-efficient, resulting in sleep modes and low-power displays, enabling users to make

50


cost versus performance and cost versus capability tradeoffs. Many proposed higher-level protocols require mobile nodes to constantly monitor access points or handshake with base stations for the purpose of synchronization, pointer control, or exchanging state information. Therefore, very limited power should be used for packet transmission. Sleep mode should be possible at the receiver front end. The active receive mode may consume more battery power than transmission mode operation since modern commercial digital communication systems may typically have transmission power of 100 mW but need 100 mA of current to support the digital signal processor operation at the receiver. 6. The maximum number of nodes and maximum coverage area: According to market studies, a WLAN may need to support hundreds of nodes. Therefore, a MAC should not limit the maximum number of nodes to maintain a satisfactory performance. This feature does not imply that we have an unlimited coverage area that is limited by the delays. The typical coverage area for WLANs ranges from 2 2 10m to 100m , which introduces less than a 1,000 ns propagation delay. WLANs are likely to operate at more than 1020M chips per second (c/s) for DS-SS and more than 1M symbols per second for other PHYs. Delay in the 5001,000 ns range can cause big problems for some MACs: a synchronous CDMA system, for instance. We can summarize this property as the ability to work in a wide range of systems, with a MAC design that can handle the geographical size and number of nodes in the LAN. 7. Robustness vs. cochannel access and interference: A big challenge to designing a WLAN MAC is to work successfully in the case of collocated networks, which can cause severe cochannel interference. It is quite likely for two or more WLANs to operate in the same region or in some regions where interference between different LANs may occur. Some protocols cannot function normally in this situation. For instance, consider two WLANs operating in two nearby buildings. For certain parts of these LANs, it may be more difficult to communicate with other parts of their own LAN than to communicate with the other LAN. Serious trouble can result from this situation if the MAC uses token passing. It is possible to mistakenly pass the token to a node in the other network. Generally


51

speaking, there are two concerns for collocated networks, described as follows: ω

ω

Security: Other users may illegally break into the network, causing a security alert. This can be solved by an appropriate authentication procedure for new users. Interference from collocated networks: For example, if we apply traditional CSMA protocols in WLANs, interference from another network can cause disastrous hidden terminal problems. These two concerns should be treated in more depth. Interference in wireless communications can be caused by simultaneous transmissions (i.e., collisions) by two or more sources sharing the same frequency band. Collisions are typically the result of multiple stations waiting for the channel to become idle and then beginning transmission at the same time. Collisions are also caused by the hidden terminal, problem, where a station, believing the channel to be idle, begins transmission without successfully detecting the presence of a transmission already in progress. Interference is also caused by multipath fading, which is characterized by random amplitude and phase fluctuations at the receiver. The reliability of the communications channel is typically measured by the average BER. For packetized voice, packet loss rates on the order of 10-2 are generally acceptable; for uncoded data, a BER of 10-5 is regarded as acceptable. Automatic repeat request (ARQ) and forward error correction (FEC) are used to increase reliability.

ω

In infrastructure LANs, multicell coverage is governed by an AP, which is typically a base station or a repeater. The coverage of each cell should overlap the neighboring cell(s) properly; that is, the overlapping region is intended to be minimized to increase system capacity but also kept to a certain proportion so that seamless service is possible. This joint region between cells introduces extra problems. These problems are summarized as follows (see Figure 3.2): ω

Self-interference: When two APs (such as two repeaters) try to transmit a packet to a node in the joint region simultaneously. This causes interference and the packet is likely to be lost.

52

Wireless LAN Standards and Applications AP-1

AP-2

AP-1

AP-2

STA (B)

STA (A) AP-1

AP-2

STA-1

(C)

STA-2

Figure 3.2 Typical cases of interference: (a) self-interference, (b) self-collision, and (c) and up-down collision.

ω

ω

Self-collision: A node in the joint region transmits a packet that is received by more than one AP. This causes collision or bandwidth waste while routing this packet to its destination(s). Up-down collision: A node (A) in one cell is transmitting uplink while another node (B) in another cell is receiving downlink. It is possible that node B may be able to hear (receive) node As transmission, and this situation results in collision, unless we can perfectly schedule all transmissions. Fortunately, this situationwhich is similar to the hidden terminal problem and is a problem as yet unsolved in multicell infrastructure LANsis very unlikely if the cells are well separated. Since up-down collisions are very destructive, any MAC should take it into account carefully.

The first two problems can be eliminated by careful coordination between APs. However, the up-down collision can only be alleviated; it can never be solved practically. A possible collision window will always exist, although it can be kept to a minimum if


53

coordination between APs and uplink/downlink takes care of this problem. The other issue is security. In a wired network, the transmission medium can be physically secured, and access to the network is easily controlled. A wireless network is more difficult to secure, since the transmission medium is open to anyone within the geographical range of the transmitter. Data privacy is usually accomplished in a radio medium using encryption. While encryption of wireless traffic can be achieved, it is usually at the expense of increased cost and decreased performance of the MAC. IEEE 802.11 supports the 802.11 draft standard, which specifies an (optional) data encryption algorithm called the wired equivalency privacy (WEP) algorithm. The WEP algorithm is based on the RC4 PRNG algorithm developed by RSA Data Security, Inc. 8. Establishing peer-to-peer connectivity without a priori knowledge: The MAC of a WLAN should support ad hoc networking. Therefore, there should be no requirement for a priori information about network topology (e.g., whether there is communication between all nodes). 9. The ability to support hand-off/roaming between service areas: At first, it was thought that a MAC protocol had to support a hand-off function to serve nodes moving from one cell to another. Currently, however, this is not a real limiting consideration because portable computers are not real mobile computers, and users normally work in a fixed place. Nevertheless, a MAC should consider this issue, which is a special feature of WLANs. In indoor environments, due to fast fading, hand-off is not a straightforward problem. For time-bounded services, the ability of a MAC to support hand-off in real time is not an easy task either, especially if we take power consumption into consideration. 10. The ability to support broadcast (multicast): Although broadcasting is the natural form of communication for downlink traffic in wireless networks, the MAC should support multicast. 11. Insensitivity to capture effects: Although the capture effect can increase throughput, it can also prohibit fair access. One solution is

54


to enforce insensitivity at the receiver end. A MAC is expected to maintain receiver sensitivity to enhance physical transmission and avoid any potential problems from capture. 12. Support of priority and non-reciprocal traffic: In addition to the timebounded services mentioned earlier, the MAC is expected to support traffic with different priorities. A special feature of WLAN traffic is that the downlink traffic is often much greater than the uplink traffic. A good MAC should definitely support this feature. Finally, although this is not strictly a MAC concern, another principal consideration should be human safety. Research is currently ongoing to determine whether RF transmissions from radio and cellular phones are linked to human illness. Networks should be designed to minimize the power transmitted by network devices. For IR WLAN systems, optical transmitters must be designed to prevent vision impairment. MAC protocols must be able to work with emission levels low enough to avoid safety complications. 3.2.2

The structure of the IEEE standard MAC protocol

An 802.11 network [14, 15] in general, consists of one or more basic service sets (BSS) that are interconnected with a distribution system (DS). BSS is the fundamental building block of the IEEE 802.11 architecture. A BSS is defined as a group of stations (STAs) that are under the direct control of a single coordination function. This coordination function can be DCF or PCF, both of which are defined below. The geographical area covered by the BSS is known as the basic service area (BSA), which is analogous to a cell in a cellular communications network. Conceptually, all stations in a BSS can communicate directly with all other stations in a BSS. However, transmission medium degradations due to multipath fading, or interference from nearby BSS reusing the same physical-layer characteristics (e.g., frequency and spreading code or hopping pattern) can cause some stations to appear hidden from other stations. An ad hoc network is the deliberate grouping of stations into a single BSS for the purpose of internetworked communications without the aid of an infrastructure network. Figure 3.3 is an illustration of the components of an IEEE 802.11 and the detail of an independent BSS (IBSS). This is the formal name of an ad hoc network in the IEEE 802.11. Any station can establish a direct communications session with any other


55 STA

STA

AP-2 IEEE 802.X Portal STA

Distribution system

STA

STA STA

(A)

Figure 3.3 networks.

AP-1

STA

(B)

Types of connectivity WLAN: (a) Ad hoc, and (b) infrastructured

station in the BSS, without the requirement of channeling all traffic through a centralized AP. We can summarize the characteristics of an ad hoc architecture as follows: ω

Lack of an AP;

ω

No functionality to support mobility;

ω

Only support data transfer between stations belonging to the same WLAN.

In contrast to the ad hoc network, infrastructure networks are established to provide wireless users with specific services and range extensions. Infrastructure networks in the context of IEEE 802.11 are established using APs. The AP is analogous to the base station in a cellular communication network. The AP supports range extensions by providing the integration points necessary for network connectivity between multiple BSSs, thus forming an extended service set (ESS). The ESS has the appearance of one large BSS to the LLC sublayer of each STA. The ESS

56


consists of multiple BSSs that are integrated together using a common DS. The DS can be thought of as a backbone network that is responsible for MAC-level transport of MAC service data units (MSDUs). The DS, as specified by IEEE 802.11, is implementation-independent. Therefore, the DS could be a wired IEEE 802.3 Ethernet LAN, an IEEE 802.3 token bus LAN, an IEEE 802.5 token ring LAN, a FDDI-MAN, or another IEEE 802.11 wireless medium. Note that while the DS could physically be the same transmission medium as the BSS, they are logically different, because the DS is solely used as a transport backbone to transfer packets between different BSSs in the ESS. An ESS can also provide gateway access for wireless users to a wired network such as the Internet. This is carried out via a device known as a portal. The portal is a logical entity that specifies the integration point in the DS where an IEEE 802.11 network integrates with a non-IEEE 802.11 network. If the network is an IEEE 802.X, the portal incorporates functions that are analogous to a bridge; that is, it provides range extensions and the transfer between different frame formats. Figure 3.3(b) illustrates a simple ESS developed with two BSSs, a DS, and a portal access to a wired LAN. Coordination functions IEEE 802.11 supports one mandatory and two optional coordination function schemes, which can be summarized as follows:

1. Distributed coordination function (DCF), based on a CSMA with collision avoidance (CSMA/CA) protocol. 2. DCF with handshakingthe request to send (RTS)-clear to send (CTS) procedureis an optional CF. The use of these two control frames limits the effect of the hidden station problem; some authors [16] call it DFW, as it uses a distribution four-way handshake. 3. Point coordination function (PCF) for distributed time-bounded services (DTBS) in which a point coordinator (or PCF station) has priority control of the medium. That is, when the PCF is active, the PCF station allows only a single station in each cell to have priority access to the medium at any one time. These CF schemes are described in more detail in the following sections.


57

Distributed coordination function (DCF) The DCF is the fundamental access method used to support asynchronous data transfer on a best effort basis. As identified in the specification, all stations must support the DCF. The DCF operates solely in the ad hoc network and either operates solely or coexists with the PCF in an infrastructure network. The MAC architecture is depicted in Figure 3.4, where it is shown that the DCF sits directly on top of the PHY and supports contention services. Contention services imply that each station with an MSDU queued for transmission must contend for access to the channel and, once the MSDU is transmitted, must recontend for access to the channel for all subsequent frames. Contention services promote fair access to the channel for all stations. The DCF is based on CSMA/CA, which is attractive to both vendors and researchers [1719] due to the popularity of Ethernet. CSMA is a member of the ALOHA family of protocols. ALOHA was the first multiple/random access protocol to be applied to large-scale wireless networks. Pure ALOHA can present unstable behavior when many collisions must be handled. This can result in an unacceptable degradation of the throughput. Improved versions (slotted ALOHA) reduce the possibility of collision duration. CSMA, which senses the status of a channel before transmitting, is the simplest way to improve ALOHA. As shown in many earlier investigations, CSMA demonstrates an increase in throughput in its various versions [20]. Therefore, the IEEE 802.3 committee chose persistent CSMA/CD as the MAC for wired LANs. The success of CSMA/CD in Ethernet relies on the ease of sensing the carrier by measuring the current or voltage in the cable. This is the primary reason why CSMA has been successfully applied in wired networks even though it was originally designed for radio networks. Despite advances in technology, carrier

PCF (For contention free services) Optional

Mandatory

DCF-RTS/CTS DCF (For contention services and basis to PCF) MAC

Figure 3.4

MAC architecture for the IEEE 802.11 standard.

DCF

58


sensing is still a major problem for radio networks due to the hidden terminal problem mentioned in Kleinrock and Tohagis pioneering paper [20]. Reliable carrier (or transmission) sensing is extremely difficult owing to severe channel fading in indoor environments and the use of directional antennas. In IEEE 802.11, carrier sensing is performed at both the air interface, referred to as physical carrier sensing, and at the MAC sublayer, referred to as virtual carrier sensing. Physical carrier sensing detects the presence of other IEEE 802.11 WLAN users by analyzing all detected packets and detects activity in the channel via relative signal strength from other sources. CSMA/CD is not used, because a station is unable to listen to the channel for collisions while transmitting. In radio systems that depend on the physical sensing of the carrier, a problem called the hidden node problem [8, 21, 22] arises. In this situation, a single receiving station can hear (i.e., is in radio range of) two different transmitters, but the two transmitters cannot hear the carrier signals of one another. In this type of topology, the transmitters send frames without performing a random backoff (because the carrier signal of the other transmitter is never heard). This results in the likelihood of collision. To alleviate the hidden terminal problem and to increase reliability, CSMA/CA is used. In general, it is associated with polling or handshaking, because multipath fading of indoor channels usually lasts for a time equal to a symbol. The 802.11 MAC protocol [14] includes, as an option, a well-known mechanism for solving the hidden node problem. The protocol makes use of two control frames: CDF with handshaking (RTS-CTS procedure)

ω

A RTS frame that a potential transmitter sends to a receiver;

ω

A CTS frame that a receiver sends in response to a transmitters RTS frame.

The CTS frame gives the requesting station permission to transmit while notifying all stations within radio range not to initiate any transmissions for a given time. This is called the net allocation vector (NAV) in 802.11. NAV indicates the amount of the time that must elapse before the current transmission session is complete and the channel can be sampled again for idle status. The channel is marked busy if either the physical or virtual carrier sensing mechanisms indicates that the channel is busy. Because of the signaling overhead involved, the RTS/CTS feature is not


59

used for short packets, for which the likelihood of collision and cost (in terms of retransmission time) are both small anyway. If RTS/CTS is not used, the duration field of the data frames actualizes NAV. K. C. Chen [6] studied the performances of this CSMA-CA with four-way handshake. We assume that the RTS-CTS transmission cycle occupies a short transmission time. The throughput presents a significant increase (10%) when a hidden terminal problem occurs. However, it pays a price in the case where there is no hidden terminal. The throughput is reduced to 63%, a substantial reduction from the original CSMA (about 80% of peak throughput); the cause is, of course, the RTSCTS overhead. To describe briefly the DCF transmission procedure, it is interesting to see Crow et al. [7] and the standard [14]. A source station performs virtual carrier sensing (NAV actualization) by sending MPDU duration information in the header of RTS, CTS, and/or data frames. An MPDU is a complete data unit that is passed from the MAC sublayer to the physical layer. The MPDU contains information, payload, and a 32-bit CRC. The duration field indicates the amount of time (in microseconds) after the end of the present frame in which the channel will be used to complete successful transmission of the data or management frame. Priority access to the wireless medium is controlled through the use of the interframe space (IFS), a time interval between the transmission of frames. The IFS intervals are mandatory periods of idleness in the transmission medium. Three IFS intervals are specified in the standard: short IFS (SIFS), point coordination function IFS (PIFS), and DCF IFS (DIFS). The SIFS interval is the smallest IFS, followed by PIFS and DIFS, respectively. Stations only required to wait a SIFS have priority access over those stations required to wait for a PIFS or DIFS before transmitting; therefore, SIFS has the highest-priority access to the communications medium. For the basic access method, when a station senses that the channel is idle, the station transmits an MPDU. The receiving station calculates the checksum and determines whether the packet was received correctly. Upon receipt of a correct packet, the receiving station waits for a SIFS interval and transmits a positive ACK frame to the source station, indicating that the transmission was successful. When the data frame is transmitted, the duration field of the frame is used to let all stations in the BSS know how long the medium will be busy. All stations hearing the data frame adjust their NAV based on the duration field value, which includes the SIFS interval and the ACK following the data frame.

60


As mentioned previously, since a source station in a BSS cannot hear its own transmissions, when a collision occurs, the source continues transmitting the complete MPDU. If the MPDU is large (e.g., 2,300 octets) a lot of channel bandwidth is wasted due to a corrupt MPDU. To reserve channel bandwidth prior to transmission of an MPDU and to minimize the amount of bandwidth wasted when collisions occur, a station can use RTS and CTS control frames. RTS and CTS control frames are relatively small (RTS is 20 octets, and CTS is 14 octets) when compared to the maximum data frame size (2,346 octets). The source station (after successfully contending for the access to the channel) first transmits the RTS control frames with a data or management frame queued for transmission to a specified destination station. All stations in the BSS hear the RTS packet with a CTS packet after a SIFS idle period has elapsed. Stations hearing the CTS packet look at the duration field and again update their NAV. Upon successful reception of the CTS, the source station is virtually assured that the medium is stable and reserved for successful transmission of the MPDU. Note that stations are capable of updating their NAVs based on the RTS from the source station and CTS from the destination station, which helps to combat the hidden terminal problem. Stations can choose from among the following options: ω

Never use RTS/CTS;

ω

Use RTS/CTS whenever the MSDU exceeds the value of RTS_ Threshold (configuration parameter);

ω

Always use RTS/CTS.

If a collision occurs with an RTS or CTS MPDU, far less bandwidth is wasted when compared to a large data MPDU. However, for a lightly loaded medium, additional delay is imposed by the overhead of the RTS/CTS frames. Figure 3.5 is a timing diagram illustrating the successful transmission of a data frame comparing both cases (using or not the RTS/CTS mechanism). Large MSDUs handed down from the LLC to the MAC may require fragmentation to increase transmission reliability. To determine whether to perform fragmentation, MPDUs are compared to the manageable parameter Fragmentation_Threshold. If the MPDU size exceeds the value of Fragmentation_Threshold, the MSDU is broken up into multiple fragments. The resulting MPDUs are of size Fragmentation_Threshold, with


DIFS

61

SIFS

DIFS

DATA

Source

SIFS

SIFS

SIFS DATA

RTS

DIFS

DIFS CTS

ACK

Dest.

ACK

CW

CW

Other

NAV

NAV NAV (RTS) NAV (CTS)

(A)

(B)

Figure 3.5 The timing diagram of a successful data frame transmission: (a) without handshaking, and (b) using the RTS/CTS mechanism.

SIFS Source

SIFS

SIFS

Fragment 0

SIFS

SIFS

Fragment 1

SIFS

DIFS

Fragment 2

CW

Dest. ACK 0

ACK 1

ACK 2

Other NAV (CTS) NAV (Fragment 0)

NAV (Fragment 2) NAV (Fragment 1)

Other NAV (ACK 0)

NAV (ACK 1)

Figure 3.6 The timing diagram of a successful fragmented data frame transmission.

the exception of the last MPDU, which is of variable size not exceeding Fragmentation_Threshold. When an MSDU is fragmented, all fragments are transmitted sequentially (Figure 3.6). The channel is not released until the complete MSDU has been transmitted successfully, or the source station fails to receive an acknowledgment for a transmitted

62


fragment. The destination station positively acknowledges each successfully received fragment by sending a DCF ACK back to the source station. The source station keeps control over the channel throughout the transmission of the MSDU by waiting only for a SIFS period after receiving an ACK and transmitting the next fragment. When an ACK is not received for a previously transmitted frame, the source station halts transmission and recontends for the channel. Upon gaining access to the channel, the source starts transmitting beginning with the last unacknowledged fragment. If RTS and CTS are used, only the first fragment is sent using the handshaking mechanism. The duration value of RTS and CTS only accounts for the transmission of the first fragment through the receipt of its ACK. Stations in the BSS thereafter maintain their NAV by extracting the duration information from all subsequent fragments. Next, the DCF collision avoidance (basic access) procedure is presented. The collision avoidance portion of CSMA/CA is carried out through a random backoff procedure. If the medium is busy, the station defers until after a DIFS is detected and then generates a random backoff period for an additional defer interval before transmitting. Referred as the contention window (CW), this minimizes collisions during contention between multiple stations. The backoff period is the unit of measurement used by the backoff timer. The backoff timer is decreased only when the medium is idle; it is frozen when the medium is busy. After a busy period the decreasing of the backoff timer resumes only after the medium has been free longer than the DIFS. A station initiates a transmission when the backoff timer reaches zero. The backoff interval is chosen following Int[22+i·ranf()]·Slot_Time, where i is the number of consecutive times a station attempts to send an MPDU, ranf() is a uniform random variable in (0,1), and Int[x] represents the largest integer less than or equal to x. ranf() is a pseudo-random number between 0 and 1, while Slot_Time is a time period. To reduce the probability of collisions, after each unsuccessful transmission the contention window takes the next value in the series until it reaches CWmax. The contention window will remain at CWmax for the remaining retries. If a station with a frame to transmit initially senses that the channel is busy, then the station waits until the channel becomes idle for a DIFS period and then computes a random backoff time. For IEEE 802.11, time is slotted in time periods that correspond to a Slot_Time. Unlike slotted ALOHA, where the slot time is equal to the transmission time of one packet, the Slot_Time used in IEEE 802.11 is much smaller than an MPDU and is used to define the IFS intervals and determine the backoff


63

time for stations in the contention period (CP). The Slot_Time is different for each physical-layer implementation. It corresponds to the sum of clear channel assessment time (time required to determine that the channel is idle), Rx_Tx turnaround time (time required for the modem to change from a receiving to transmitting configuration and vice versa), and air propagation time. The contention window takes an initial value CWmin for each frame queued for transmission. The advantage of this channel access method is that it promotes fairness between stations, but its weakness is that it probably could not support DTBS. Fairness is maintained because each station must recontend for the channel after every transmission of an MSDU. All stations have equal probability of gaining access to the channel after each DIFS interval. Time-bounded services typically support applications that must be maintained with a specified minimum delay, such as packetized voice or video. With DCF, there is no mechanism to guarantee minimum delay to stations supporting time-bounded services [23, 24]. Point coordination function (PCF) The PCF is an optional capability that is connection-oriented and provides contention-free frame transfer. The PCF relies on the point coordinator (PC) to perform polling, enabling polled stations to transmit without contending for the channel. The AP within each BSS performs the function of the point coordinator. Stations within the BSS that are capable of operating in the contention-free period (CFP) are known as CF-aware stations. The method by which the polling tables are maintained and the polling sequence is determined is left to the implementers. The main applications of this capability are time-bounded services (usually packetized voice or video) in which packet order is a main concern. The PCF is required to coexist with the DCF and logically sits on top of the DCF (see Figure 3.4). The CFP repetition interval (CFP_Rate) is used to determine the frequency with which the PCF occurs. Within a repetition interval, a portion of the time is allotted to contention-free traffic, and the remainder is provided for contention-based traffic. A beacon frame initiates the CFP repetition interval, where the AP transmits the beacon frame. One of its primary functions is synchronization and timing. The duration of the CFP repetition interval is a manageable parameter that is always an integer number of beacon frames. Once the CFP_Rate is established, the duration of the CFP is determined. The maximum size of the CFP is determined by the manageable parameter CFP_Max_Duration. It varies between a minimum (time required to

64


transmit two maximum-sized MPDUs, including overhead, the initial beacon frame, and a CF-End frame) and a maximum (CFP repetition interval minus the time required to successfully transmit a maximumsized MPDU during the CP, including time for RTS/CTS handshaking and ACK). Therefore, time must be allotted for at least one MPDU to be transmitted during the CP. It is up to the AP to determine how long to operate the CFP during any given repetition interval. If traffic is very light, the AP may shorten the CFP and provide the remainder of the repetition interval for the DCF. The CFP may also be shortened if DCF traffic from the previous repetition interval carries over into the current interval. The maximum amount of delay for a frame or fragment on the CFP interval is the time needed to transmit an RTS/CTS handshake, a maximum length MPDU, and ACK. Figure 3.7 shows the CFP repetition interval, illustrating the coexistence of the PCF and DCF. At the nominal beginning of each CFP repetition interval, all stations in the BSS update their NAV to the maximum length of the CFP (i.e., CFP_Max_Duration). During the CFP, the only time stations are permitted to transmit is in response to a poll from the point coordinator or for transmission of an ACK (a SIFS interval after receipt of an MPDU). Now, the point coordination funciton (PCF) transmission procedure is presented. At the nominal start of the CFP, the point coordinator senses the medium. If the medium remains idle for a PIFS interval, the point coordinator transmits a beacon frame to initiate the CFP. The point

CFP Repetition interval

CFP Repetition interval

Beacon frames PCF

CFP

DCF

CP

NAV

Figure. 3.7

PCF

CFP repetition interval.

NAV

DCF


65

coordinator starts contention-free (CF) transmission a SIFS interval after the beacon frame is transmitted by sending a CF-Poll (no data), Data, or Data+CF-Poll frame. The point coordinator can immediately terminate the CFP by transmitting a CF-End frame, which is common if the network is lightly loaded and the point coordinator has no traffic buffered. If a CF-aware station receives a CF-Poll (no data) frame from the point coordinator, the STA can respond to the point coordinator after a SIFS idle period, with a CF-ACK (no data) or a Data+CF-ACK frame. If the point coordinator receives a Data+CF- ACK frame from a station, the point coordinator can send a Data+ CF-ACK+CF-Poll frame to a different station, where the CF-ACK portion of the frame is used to acknowledge receipt of the previous data frame. The ability to combine polling and acknowledgment frames with data frames, transmitted between stations and the point coordinator, was designed to improve efficiency. If the point coordinator transmits a CF-Poll (no data) frame and the destination station does not have a data frame to transmit, the station sends a nullfunction (no data) frame back to the point coordinator. Figure 3.8 illustrates both frame transmission cases, that between the point coordinator and a station and station-to-station. If the point coordinator fails to receive an ACK for a transmitted data frame, the point coordinator waits a PIFS interval and continues transmitting to the next station in the polling list. After receiving the poll from the point coordinator, the station may choose to transmit a frame to another station in the BSS. When the destination station receives the frame, a DCF ACK is returned to the source station, and the point coordinator waits a PIFS interval following the ACK frame before transmitting any additional frames. The point coordinator may also choose to transmit a frame to a non-CF-aware STA. Upon successful receipt of the frame, the STA would wait a SIFS interval and reply to the point coordinator with a standard contention-period ACK frame. Fragmentation and reassembly are also accommodated within the Fragmentation_Threshold value used to determine whether MSDUs are fragmented prior to transmission. It is the responsibility of the destination station to reassemble the fragments to form the original MSDU. 3.2.3

Comparison with the MAC protocol of other WLANs: HIPERLAN

This section compares the performances of the MAC protocol with the protocol of its main potential commercial competitor: HIPERLAN. HIPERLAN is extensively described in Chapter 4; here we summarize the most important characteristics of this standard.

D1+Poll

SIFS

NAV

PIFS

(B)

U2+ACK

SIFS

D2+ Poll

SIFS CF-End

D4+Poll

SIFS

C. P.

U4+ACK

CF-End

SIFS

C. P.

Figure 3.8 The timing diagram of the frame transmission (a) between the point coordinator and a station, and (b) station-to-station in PCF.

SIFS

ACK

Contention-free period

U1+ACK

(A)

NAV

PIFS

D3+ACK+Poll

CFP repetition interval

SIFS

U2+ACK

SIFS

Contention-free period

D2+ACK+Poll

SIFS

U1+ACK

SIFS

D1+ Poll

SIFS

Beac

PIFS

PIFS

Beac

SIFS

66 Wireless LAN Standards and Applications


67

The HIPERLAN MAC protocol [21, 2528] is based on a carriersensing mechanism but is quite different in its details from that used in the IEEE 802.3 standard (Ethernet) or the IEEE 802.11 standard. If the medium has been sensed as free for a sufficient length of time, 1,700 bit times in this case, immediate transmission is allowed. If not, the channel access, in the terminology used in the HIPERLAN standard, consists of three phases: prioritization, elimination, and yield. The actions of each node in these three phases are described below and in Figure 3.9. The prioritization phase aims at allowing only those nodes having packets of the highest available priority to contend further for channel access. This phase consists of a number of slots, with a node having a packet with priority p transmitting a burst1 in slot p + 1 if it has heard no higher-priority burst. At the end of the first burst on the channel, the prioritization phase ends and the elimination phase begins. During the elimination phase, nodes that transmitted a burst during the prioritization phase now contend for the channel. This is achieved by each node transmitting a burst for a geometrically distributed number of slots and then listening to the channel for one time slot. If another burst is heard while listening to the channel, the node stops contending for the channel. Thus, only the node(s) with the longest burst will, in the absence of the hidden node problem, be allowed to further contend for the channel. Immediately after the longest burst and listening period of the elimination phase, the yield phase is started. In this phase, each of the surviving

1-5 slots

Prioritization

n+1slots

m slots

n≤12

m≤15

Elimination

Yield

Transmission ends Figure 3.9

Transmission

Three phases of HIPERLAN channel access.

1. Roughly speaking, a burst entails transmitting the carrier frequency. More precisely, there is a particular bit sequence that is repeated for the duration of a burst, but all receivers respond only to the received signal strength and not to the particular bit sequence.

68


nodes defers transmission for a geometrically distributed number of slots, while listening to the channel. However, if they hear any transmission, they defer transmission altogether. The purpose of the elimination phase is to bring the number of contenders down to small number, and then the yield phase tries to ensure that only one node eventually transmits. As a result, the chances of actual collisions for data are negligible (less than 3%). The HIPERLAN technical committee explicitly wanted to support a QoS for packet delivery. QoS support is provided via two mechanisms: the initial value in both cases being assigned by the application using the HIPERLAN services and the priority of a packet (high or normal) and the packet lifetime measured in integral milliseconds with a range of 0-32,767 ms (default value: 500 ms). The residual lifetime of a packet together with its priority are used to determine its channel access priority. Channel access priority can fall into one of five categories and is used for the prioritization phase described above. No other explicit mechanism is used to support the desired QoS, unlike the time-bounded services of the IEEE 802.11 standard. The committee envisioned that a pure cellular architecture would not be sufficient for the system, hence allowing HIPERLAN nodes to forward packets destined for other nodes. This, of course, requires the maintenance of routing databases at nodes and the dynamic updating of these databases. Methods for this topology maintenance have also been addressed in the standard, for both the databases at each node and broadcasting the information to other nodes. However, it is optional for a node to forward packets; hence, a node can also choose to forego this function, becoming a nonforwarder in the terminology. An interesting discussion of some of the issues involved in this process can be found in [25]. There is support for packet encryption in the HIPERLAN packet transmission mechanism. The standard stays away from defining the particular encryption method used but defines methods to inform the receiver which of a particular set of encryption keys has been used to encrypt the packet. The standard defines a small set of such keys and how they are kept at nodes. It does not, however, define any key distribution strategy, which would be a management function on top of the basic services. Another ETSI committee is working on a security standard for HIPERLAN that will be required for conformance. Table 3.1 summarizes and compares the main characteristics of the HIPERLAN 1 and basic IEEE 802.11 protocols.


69

Table 3.1 A Comparison of the Main Characteristics of the HIPERLAN 1 and IEEE 802.11 Protocols HIPERLAN 1 MAC IEEE 802.11 MAC

3.2.4

Allows ad hoc networks

Yes

Yes

Supports AP (bridge to wired networks)

Yes

Yes

Power-saving mode

Yes

Yes

Encryption

Yes

Yes

Authentication

No

Yes

Association

No

Yes

QOS

Yes

Nonstandardized

Conclusions

Some conclusions can be obtained on the IEEE 802.11 MAC structure and performance. First, studies on efficiency [7, 23, 24] show that DCF is superior to PCF for short MSDU and low values of BER. On the other hand, if MSDU length increases, PCF performances are superior, as in packetized services (such as voice or video). The 802.11 standard MAC structure is a compromise between several proposals. It means that MAC has to cooperate with different PHY layers and leave decision-making capabilities to implementers. We believe that some optional capabilities (especially handshaking in DCF) will become mandatory in standards.

3.3 Physical layer for IEEE 802.11 wireless LANs: Radio systems 3.3.1

Introduction

There are two specifications for radio systems within the IEEE 802.11 physical-layer definition: the FH-SS physical layer and the DS-SS physical layer. Both use spread spectrum techniques and employ radio transmission in unlicensed spectrum bands at around 2.4 GHz. Unlicensed frequency bands have increased in interest and importance over the last decade because of the technological advances that have allowed the development of compact and cheap radio transceivers. Many applications (preexistent or new) can be implemented using these transceivers to communicate data of a different kind.

70


Traditionally, radio spectrum has been considered a scarce natural resource whose use has to be regulated by national administrations. These administrations decide on the users that are granted a license for spectrum use and impose some kind of payment for this use. This scheme works well for traditional users (public operators, broadcasters, and government agencies) but cannot be efficiently applied when the number of potential users of a particular application could number millions. This is not only the case for WLANs but also for other applications such as the ubiquitous garage door opener. Unlicensed bands, such as the ones known as ISM (industrial, scientific, and medical) bands, can be employed by anyone using equipment that complies with some technical specifications. The regulation applies to the transceivers that have to be designed to meet some characteristics usually to limit the amount of interference with other users. As there is no frequency planning, some degree of interference cannot be avoided. The user can freely employ the equipment. There is no possibility of claiming protection from interference caused by other users. Potential problems must be solved by private agreements. To reduce the probability of interference, transmission power is strictly limited within these bands, in many cases at levels as low as 10 mW or 100 mW. This limitation guarantees that the potential interference sources are physically in the vicinity of the receiver. This is sufficient for many applications. For example, within industrial buildings, it is very unlikely that external transmitters can interfere with the factory communications systems. The same can be said of systems inside hospitals as well as for most indoor systems. Outdoor systems may suffer from interference from transmitters situated in the vicinity, but many of them, like the aforementioned garage opener, are also protected by the discontinuous nature of the transmission. In 1985, the FCC decided to augment the transmission level in some unlicensed bands to 1W to boost the development of new applications requiring a greater range. To limit the amount of potential interference, the agency established tight limits in the transmission power spectral density, so that the use of spread spectrum systems was promoted: The greater the band, the greater the transmission power. The new rules, known as Part 15, included regulations for direct sequence (DS) and frequency hopping (FH) systems. The 802.11 PHY radio systems can be seen, in a sense, as a consequence of this regulation. It is adapted to the FCC specifications for the 2.4-GHz ISM band. As many countries have adopted similar regulations,


71

the standard can be used worldwide. Some limitations are included in the standard to take into account differences in the regulations of particular countries. Thus, there are technical and regulatory issues behind the specifications of physical layer spread spectrum systems. Regulatory issues have also limited the available band and bit rate. It is foreseen that new specifications for higher bit rates will be developed within the standard, according to regulations for other unlicensed bands. Section 3.3.2 discusses the technical characteristics of spread spectrum systems and their potential advantages. In addition, Section 3.3.2 reviews the fundamentals of FH and DS transmission systems and describes the main specifications of the 802.11 PHY, briefly noting the advantages and problems of each of the systems. Finally, a comparison of the two systems is made. 3.3.2

Spread spectrum techniques

Spread spectrum systems use more bandwidth than that needed for transmission. To determine how much bandwidth is needed for a transmission, it is necessary to consider the modulation format. Many people would agree that a system could be considered spread spectrum if the occupied bandwidth is intentionally made greater than that needed, given the bit rate and the modulation format. If the system uses roughly the bandwidth needed then it is not a spread spectrum system and should be considered to be narrowband. (Narrowband will be used in this context as the opposite of spread spectrum.) To achieve a larger bandwidth, spread spectrum systems use a code in the transmitter, independent of the data, prior to modulation, that must be known by the receiver. A receiver unaware of the code would be unable to decode the transmitted data. When compared to narrowband transmissions, spread spectrum transmissions are more difficult to detect, intercept, or decode. Thus the main applications of these systems were initially military. However, the same properties have advantages in commercial systems as they are less sensitive to interference from other users and less likely to interfere with others. This is particularly true when the interfering/interfered user uses narrowband transmission. Both kinds of systems can coexist in the same frequency band with little disturbance to each other. There are no differences between narrowband and spread spectrum systems in noisy environments. Their performance in additive white

72


Gaussian noise (AWGN) channels is exactly the same; thus, their range is equal. Spread spectrum systems are in general more complex and thus have been adopted to commercial systems, only when technological advances have allowed the integration of powerful digital signal processors that can be manufactured in large quantities at very low costs. The most frequently used spread spectrum techniques are FH and DS. As both have been included in the IEEE 802.11 standard, Sections 3.3.3 and 3.3.4 present their main features. More detailed descriptions of spread spectrum systems and their applications can be found in [2932]. 3.3.3

Frequency hopping techniques

FH systems use conventional modulation techniques, but the carrier frequency is changed at a given rate, following a given sequence. This sequence is the code of these systems. A receiver that does not know the code cannot follow the frequency hops and can only occasionally detect some data. If the hopping rate is faster than the bit rate, the system is fast FH (FFH). If the hopping rate is slower than the bit rate, the system is slow FH (SFH). Commercial systems are always SFH because of the complexity of FFH systems. In an SFH system the bit stream is split into packets, which are each transmitted in a burst with a different carrier frequency. Within a given burst, the transmission is narrowband, using just the bandwidth needed according to the bit rate and the modulation format. Figure 3.10 shows an example of this transmission. Disregarding for the moment the problem of generating the frequency hops in transmitter and receiver, and synchronizing them, it is obvious that there are no basic differences in performance in noisy environments compared to narrowband systems. In fact, the transmission is narrowband itself. Where are the advantages in FH systems to be found? In an interference environment, there are two, described as follows: ω

If there is a narrowband interfering transmitter, the interference affects only those bursts whose carrier coincides with the carrier of the other transmitter. On the other hand, the interference with other narrowband systems is mitigated by the fact that the transmission does not always occupy the same bandwidth. The interference only occurs at times.


73

Frequency

Occupied bandwidth Figure 3.10

ω

Time

FH transmission.

If two FH systems working on the same band have overlapping coverage areas, interference will occur every time the bursts of both systems coincide in the same carrier.

In addition, there is a third advantage that is very important in multipath environments. When the signal arrives at the receiver through different paths, with different delays, frequency-selective fading occurs: Some frequencies become heavily attenuated because of the destructive combination of the signals coming from different paths. As the resulting phase differences depend on the carrier frequency, for other frequencies the combination is not destructive and can even be constructive, reinforcing the received power. In an FH system some bursts would suffer from fading, but others would be received perfectly. In a multipath propagation environment, only some bursts are affected by frequencyselective fading. As a consequence, these systems usually assume that some of the bursts are received with very low quality (high BER) or are completely lost because of propagation impairments or because of a powerful interfering signal. However, this affects only a small percentage of the transmission and can be recovered, either by forward error coding or by retransmissions. FH systems are usually considered to work on a pass-fail basis. Some bursts are received perfectly, and others are lost. The system

74


works as long as the rate of lost bursts is small. This can be achieved in two ways: ω

Fight against interference. Usually the higher the number of available frequencies, the smaller the number of lost bursts. This is true if the spectrum is not greatly used.

ω

However, to limit the effects of selective fading, it is not enough just to use the highest possible number of frequencies. It is also necessary to limit the probability that many of them are simultaneously affected by fades. This leads to the concept of the channel correlation bandwidth, or the amount of frequency displacement for which there is a high probability that fades occur at the same time (in some sense, it is the bandwidth of fades). The FH system will tolerate selective fading with few problems if the system bandwidth is much larger than the channel correlation bandwidth.

The coherence correlation bandwidth is roughly inversely proportional to the rms delay spread, which is a measure of the difference between the time of arrival of the first and the last significant components of the signal. An estimation of the coherence bandwidth Bc makes it equal to 1/(2πD), where D is the root mean squared (rms) delay spread. Thus, it can be very small in outdoor mountainous areas, with different rays traveling very different distances. However, it is very large in indoor environments, where the distance differences are measured in meters and delay spreads in nanoseconds. In median rooms rms delay spread has been measured to be in the range of 2550 ns [33]. The coherence bandwidth would be between 3 and 6 MHz. Larger rooms would show larger delay spreads and smaller coherence bandwidths. It is clear that FH systems are not appropriate when the spectrum is heavily used. It requires that most of the bursts be properly received, free of interference, meaning that most of the channels must be free. However, this is not a problem in unlicensed bands. It is important to keep in mind that as there is neither frequency planning nor user control, unlicensed frequency bands are never used as much as licensed ones, no matter which transmission system is selected. When frequency use is more important, as in licensed bands, orthogonal FH can be used. This is a technique that allows a group of transmitters to use the same frequency band, all of them with FH, but with their hop sequences tightly synchronized, so that it is impossible for


75

two transmitters to use the same frequency at the same time. This is used, for example, with the same cell transmitters in the cellular system GSM. In a scenario of two independent FH transmitters that are working in unlicensed bands, the interference between them will occur when their frequencies coincide. On average 1/N of the bursts would be lost, N being the number of available frequencies. Actually, as the hop times would not be synchronized, the number of lost bursts would be between 1/N and 2/N on average. In any case, it is clear that the greater the number of frequencies, the higher the level of protection against interference. As N increases, the bandwidth also increases. In fact, N can be regarded as the ratio between the bandwidths for the spread spectrum system Bss and the narrowband one Bnb. N=

B ss Bnb

(3.1)

Because it is possible to combine the modulation with the hopping sequence, FH systems usually (but not necessarily) employ frequency modulation, normally some variant of FSK with two or four levels. BPSK and QPSK have also been used. FH transmitters and receivers are technologically complex as they have to rapidly change the carrier frequency. The switching time is lost for transmission and thus has to be made as small as possible. Also, it has to be very small compared to the length of the burst. Currently, commercial PLL frequency synthesizers can change frequency on the order of 100 µs, and thus the length of the bursts can be of only a few milliseconds. Hop rates in commercial systems usually are between 1 and 300 hops per second. The receivers have to synchronize their frequency hop sequence with the one embedded in the received signal. Unless some kind of centralized control informs all the nodes of the times to hop, the receiver must usually spend a considerable amount of time (one or several cycles of the hopping sequence) to acquire synchronism. There are several techniques for acquiring and keeping the hop synchronism, and its applicability depends on the hop rate and the modulation format. A detailed description of some of these techniques can be found in [29, 31]. 3.3.4

Direct sequence systems

In DS systems the modulation rate is intentionally increased to spread the spectrum. This is achieved by combining the bit sequence with a higherrate binary sequence (called the chip sequence) to obtain a new sequence

76


with the chip rate. This one is then used to modulate the carrier. Figure 3.11 shows a basic scheme for a DS transmitter. The inverse operation is performed at the receiver side. The signal is demodulated, and then it is recombined with the same chip sequence to restore the original data. Figure 3.12 shows a much simplified diagram of a DS receiver. Although these are the basic ideas behind the concept of DS systems, practical implementations require further considerations on the kind of chip sequences used, ways to combine them with the data sequence,

Data MOD

Power amplifier

Code

Carrier generator

Figure 3.11

Basic direct sequence transmitter.

DEMOD and DECODE

X Intermediate frequency amplifier

Low-noise amplifier Local oscillator

Figure 3.12

Code

Simplified diagram of a direct sequence receiver.

Data


77

modulation formats employed, techniques to demodulate the signal, synchronization issues, and data recoverers. This section addresses all of these aspects. The advantages of these systems compared to narrowband systems are described as follows: ω

The power spectral density, measured in power per unit of bandwidth, is much lower because of the larger bandwidth over which the power is spread. This influences the low probability of interception and the lower capacity of interfering with other systems.

ω

To recover the data, the receiver must know the chip sequence and carry out the operation of combining the received sequence with the chip sequence. This adds privacy to the communication, because any unwelcome listener would be unable to recover the bit sequence if the chip sequence is kept private.

ω

In the receiver, the operation to combine the signal with the chip sequence restores the data to their original bit rate, which is much smaller than the chip rate. As a consequence, the signal bandwidth is reduced, and the components lying outside this smaller bandwidth can be filtered out. As any other signals, including noise and interference of any kind, will have a wide bandwidth after their combination with the chip sequence, most of their power will be filtered. Note that narrowband interference would be spread throughout the band after combination with the chip sequence. Thus, DS receivers present a certain degree of noise rejection and interference.

ω

Several communications can share the same spectrum with direct sequence communications, as long as they use an uncorrelated chip sequence. This is the base of CDMA, a multiple access system in which each user is assigned a spreading code, so he/she is unable to interfere with one another, or mutual interference can be kept at controlled low levels.

Not all of these advantages are used in the 802.11 standard. In particular, it is not a CDMA system. The DS method does not play a role in the privacy features of this standard, as the chip sequence is public and equal for all users. The main reason for selecting this method of transmission is its ability to share the spectrum with other systems at low levels of mutual interference.

78


Codes in direct sequence systems The selection of codes is one of the most interesting aspects of designing a DS system. Codes are always periodic, but their period can be as short as a symbol period, or as long as several thousands of symbols or even more. Some codes used in the cellular system IS-95 have a period of almost one century. Short codes facilitate the acquisition of synchronism in the receiver, while long codes are needed to guarantee the privacy of the communications. There are two characteristics that are met by all the codes used in DS systems: ω

Balanced codes, with roughly the same number of 0s and 1s. These are needed to avoid the presence of a DC component. Pseudorandom sequences with roughly equal probabilities of appearance of both symbols (0s and 1s) are very often used, particularly in long codes.

ω

Because the chip rate is always a multiple of the bit rate, in a bit period there is an integer number of chips. Both sequences are synchronized. This reduces the amount of high-frequency energy that would appear if a transition in the bit sequence were followed shortly afterward by a transition in the chip sequence. The ratio of the chip rate (Vc) over the bit rate (Vb) is called the processing gain (Gp); it influences all the systems performance. It is also equal to the ratio of the spread spectrum system bandwidth (Bss) to the bandwidth of a narrowband system (Bnb) with the same bit rate. Gp =

Vc B = ss Vb Bnb

(3.2)

Codes can be classified as orthogonal, quasi-orthogonal, and uncorrelated depending on their behavior in relation to other codes of the same kind. This classification can also be applied to systems, according to the type of codes used. Orthogonal codes are those whose correlation with codes of the same family is exactly zero. This is a very good property, as it means that different users using these codes can share the same frequency band with zero interference among them. Unfortunately, this property is only met when the codes are combined with the same time reference. Even small displacements between the time origin of codes usually lead to an uncontrolled amount of interference. This means that the system must maintain perfect synchronism at the chip level among all users, which is not easy to achieve. Orthogonal codes are usually employed for the


79

transmission from a central node to different remote receivers. As the signals are generated in the same transmitter, there is no difficulty in synchronizing them at chip level. One example of orthogonal codes is the Walsh sequences used in the IS-95 cellular system [32]. Quasi-orthogonal codes are those that, while keeping a very small level of correlation when combined in phase, can tolerate displacements to some degree with no significant degradation in their mutual interference. They are the preferred choice when there is a synchronization system that can maintain alignment of the received signals during a few chip intervals. Some subsets of Gold codes have this property and are used in CDMA satellite communications [34]. Uncorrelated codes present low levels of correlation with one another, for any time displacement. They are used when no time synchronization between different users is performed by the system. Different users employ different codes, and the interference with each other is kept small on average, with only occasional peaks that would lead to bit errors. Examples of uncorrelated codes are maximal length codes and Gold codes [31]. Orthogonal and quasi-orthogonal codes are usually short codes, as this facilitates the synchronization of all the signals. They are obtained from tables in the transmitter and receiver. On the contrary, uncorrelated codes are usually long codes. This helps to limit the level of correlation between different codes and to enlarge the time between interference peaks. They are usually obtained from pseudorandom generators, implemented with displacement registers with some linear feedback. This technique is the same used in data scramblers, or in cryptography systems. Different versions of the same code, obtained from a given structure but with different initialization words, are also employed for different users, as they feature low correlation. Gold codes use two pseudorandom sequence generators, whose outputs are combined. The selection of two words for initialization allows for the generation of completely different codes, with low correlation properties but with the same generation structure. Transmission and reception in direct sequence systems The combination of the chip sequence with the bit sequence is usually made by addition (mod. 2) with an XOR gate. Several modulation formats can be used. One possibility is an I-Q modulation, such as BPSK or QPSK. QPSK has the advantage that it spreads the power further between the two orthogonal carriers with the same frequency (sine and cosine components). This

80


improves the performance of the system in the presence of narrowband interference signals. BPSK is easier to implement and has a higher immunity to noise. Some systems use BPSK for the data and QPSK for the spreading sequences to combine the advantages of both modulation formats. DS transmitters are not too different from the basic one shown in Figure 3.12. Practical implementation can vary depending on how much processing is made in digital form and how much is analog processing and whether the modulation is made on the RF frequency or in an intermediate frequency. In this case a frequency converter that transforms the signal to RF and delivers it to the power amplifier follows the modulator. The transmitter structure is quite similar to a digital transmitter working at the chip rate, whose data input signal is the combination of data and chip sequences. However, DS receivers are not at all digital receivers working at the chip rate, followed by a despread module. In fact, individual chips need not be detected in a DS receiver, and a receiver designed to detect them would present a degraded threshold, because it would have to cope with all the noise and perturbations present in the spread spectrum bandwidth. A more realistic structure of a DS receiver is shown in Figure 3.13. The signal is first down-converted and pass-band filtered within its RF bandwidth. This is the spread spectrum bandwidth Bss. The filtered signal is then XOR-combined with the code. Subsequently, it is filtered with the narrowband bandwidth Bnw. As this is smaller than Bss in the processing gain Gp, noise and perturbations are rejected in this proportion. After this filter, the signal is conceptually identical to a signal received in a narrowband system; it undergoes the usual operations of carrier and clock synchronization and detection. What is specific to a direct sequence receiver is the need for synchronization of the code with the received signal. This is performed with a local code generator controlled by a local clock. This is in practice a voltage controlled oscillator (VCO). When the code is in phase with the one used in the transmitted signals, higher levels are detected in the correlator output. The power detected at this point is filtered and fed back to control the VCO frequency. The structure is very much like a phase lock loop, but as the controlled variable is the delay between the transmission and reception codes, it is called a delay lock loop (DLL). There are several variants of this scheme. The basic delay lock loop is shown in Figure 3.14. Further reading on delay lock loops can be found in [29, 31]. Strategies for combining the operations of code, carrier, and clock synchronization


X

X Low-noise amplifier

81

Data sample and decision

Detection

Intermediate frequency Correlation amplifier

Local oscillator

Channel Automatic selection gain control

Code generator

Carrier generator

Clock generator

Delay lock loop

Carrier synchronization loop

Clock synchronization loop

Processing Unit

Figure 3.13

Typical structure of a direct sequence receiver.

X

X2 + +

F(s) -

X

X

P(t+k) P(t-k) P(t) Figure 3.14

Delay lock loop.

Code generator

2

VCO

82


as well as the automatic gain control are the keys in designing DS receivers. The process of filtering the signal within the spread spectrum bandwidth, combining it with the code, and further filtering within the narrowband bandwidth affects all the signals present in the system, including the noise, by reducing its power by roughly the process gain. It can be shown that: Performance in direct sequence systems

 S  S   = Gp ×    N  nb  N  ss

(3.3)

However, for most of the interferences, I, the (S/I)nb is:  S  S   ≈ Gp ×    I  nb  I  ss

(3.4)

The approximation becomes equality for interferences having a low degree of correlation with the signal. This includes other users interference if they use low correlation codes but can be applied to almost any other signal that can be present in the signal bandwidth. Another interesting property is that by applying the central limit theorem, it can be shown that any interfering signal, and the sum of all of them, has an amplitude distribution that is approximately Gaussian after the correlation process. The combination of all the interfering signals and noise after the correlation is noise-like. This is the optimum perturbation from the information theory [35] and, as a result, the protection against interference is maximum. The performance of a DS receiver in a particular environment can be easily analyzed, in a first approximation, by obtaining the total perturbation power within the spread spectrum bandwidth as: Pss = ∑ I i + N ss

(3.5)

and then finding the signal to perturbation ratio in the narrowband bandwidth as:  S  S   ≈ Gp ×    P  nb  P  ss

(3.6)

Then, because the sum of perturbations after the correlation process is noise-like, the relationship that, for the modulation format, relates the BER to the signal-to-noise ratio in the detector can be applied. Although


83

this method is only approximate and should be applied carefully, it allows us to remark on the main advantage of DS transmission: It allows extending the threshold as much as the processing gain. For example, a system whose threshold is 6 dB, with a processing gain of 10 dB can work with perturbations 4 dB more powerful than the received signal. Thus, signal-to-noise ratio can be −4 dB at the receiver input. However, this is not an advantage, as the broader bandwidth implicitly has a higher noise power in the receiver input. For this reason, this DS is transparent with regard to the system range. On the other hand, this is an advantage with regard to interference, as the system can tolerate 10 dB (in this example) more interference power than a narrowband system. Next, the physical layer for PH is presented. 3.3.5

IEEE 802.11 frequency hopping physical layer

Transmission in the FHSS physical layer is defined in two modes: with bit rates of 1 Mbps and 2 Mbps. In both cases, the band between 2.4 and 2.5 GHz is structured in channels. The carrier frequency for channel n can be calculated as: Radio transmission

f n = 2400 + n

(3.7)

MHz

Because of different regulations in different geographical areas, North America and most of Europe allows the use of 79 channels, but Japan, Spain, and France can use only about 30 channels. The range of frequencies is presented in Table 3.2. Channels 2 to 80 are thus available in North America and most of Europe. In Japan, channels are from 73 to 95. In France, channels range from 47 to 73, and in Spain they range from 48 to 82.

Table 3.2 Frequency Bands Available in Different Countries Geographic Region

Band Limits

Channels

North America and Europe

2.4022.480 GHz

280

France

2.4482.482 GHz

4882

Spain

2.4472.473 GHz

4773

Japan

2.4732.495 GHz

7395

84


FH sequences are defined in the standard [14] in the form of patterns that are permutations of all the frequencies available in the particular geographical area. The number of different patterns is equal to the number of channels, shown in Table 3.1. Every pattern uses all the available channels in a period. Patterns are described with a parameter x, according to the following relationships: fx(i) = [b(i) + x] mod(79) + 2 in North America and Europe, fx(i) = [(i − 1)* x] mod(23) + 73 in Japan, fx(i) = [b(i) + x] mod(27) + 47 in Spain, and fx(i) = [b(i) + x] mod(35) + 48 in France, where b(i) are the base-hopping sequences, shown in Tables 3.33.5. In North America and Europe, the minimum hop size is 6 MHz; it is 5 MHz in Japan. This is to minimize the correlation between channels used in subsequent bursts. Three sets are defined for each geographical region. Each set is made up of 26 patterns in North America, 4 patterns in Japan, 9 in Spain, and 11 in France. Within each set, long periods of collisions are avoided. Table 3.3 Base-Hopping Sequence for North America and Europe 0

23

62

8

43

16

71

47

19

61

76

29

59

22

52

63

26

77

31

2

18

11

36

72

54

69

21

3

37

10

34

66

7

68

75

4

60

27

12

25

14

57

41

74

32

70

9

58

78

45

20

73

64

39

13

33

65

50

56

42

48

15

5

17

6

67

49

40

1

28

55

35

53

24

44

51

38

30

46


85

Table 3.4 Base-Hopping Sequence for Spain 13

4

24

18

5

12

3

10

25

19

8

23

15

22

9

21

0

6

14

1

20

7

16

2

11

17

26

Table 3.5 Base-Hopping Sequence for France 17

5

18

32

23

7

16

4

33

26

10

31

20

29

22

12

13 6

28

14

25

0

8

1

15

3

11

30

24

9

27

19

2

21

34

Thus, network overlapping on the same area can use patterns of the same set to minimize the interaction. The modulation employed is Gaussian frequency shift keying (GFSK) for the two data rates. In both cases the symbol rate is 1 Msymbol/s. Two levels of GFSK are employed for 1 Mbps, and four levels of GFSK are used to transmit 2 Mbps. The Gaussian filter bandwidth multiplied by the symbol period gives BT = 0.5 for both speeds. The nominal frequency deviations are shown in Table 3.6. For 2-GFSK, h2 = 0.32. The nominal frequency deviation from the carrier is 160 KHz. This is the maximum deviation measured after a Table 3.6 Frequency Deviation for 1 Mbit/s and 2 Mbit/s.

Modulation

Symbol

Carrier Deviation (MHz)

2-GFSK (1 Mbps)

1

1/2*h2

0

−1/2*h2

10

3/2*h4

11

1/2*h4

01

−1/2*h4

00

−3/2*h4

4-GFSK (2 Mbps)

86


certain number of consecutive ones or zeroes to eliminate the effect of the Gaussian filter. For 4-GFSK, h4 = 0.144. Nominal frequency deviation for the external symbols is 216 KHz. These are nominal values. Maximums are defined by national regulations. Channel switching, defined as the time to settle to within 60 KHz of a new channel nominal frequency, is established at 224 s. For transmission efficiency, a minimum dwell time of 2 ms (maximum hop rate of 500 hop/second) would probably be used in practice. Maximum dwell time is defined by regional regulations. In the FCC it is defined as 400 ms, which gives a minimum hop rate of 1.25 hops/second. Transmit center frequency accuracy is established at 660 KHz. All receivers must be able to work with an input signal whose center frequency is in this range from the nominal channel frequency. Transmitter power must be stabilized within 2 dB of its nominal value for the time of whole frame. The maximum time to switch from on to off and from off to on is 8 µs. The time to switch from reception to transmission is 19 µs. Reference receiver sensitivity is set at −80 dBm for the 1 Mbps and −75 dBm for the 2 Mbps receiver. The threshold is established at a frame error ratio (FER) of 3% for the MPDUs of 400 octets. The maximum received signal level is −20 dBm for the correct recovery of the data. The receiver threshold for the clear channel assessment (CCA) is fixed at −85 dBm for the preamble and −65 dBm for the data for an 802.11-compliant signal. This is for transmission power lower than or equal to 100 mW. A device with transmission power greater than 100 mW should have these thresholds decreased in 5*log (Pt /100 mW). Transmitter power (Pt) depends on national regulations and implementations. The maximum transmitter equivalent radiated isotropic power (EIRP) should exceed 10 mW, except when forbidden by national regulations. Power control is mandatory if the maximum EIRP exceeds 100 mW, with at least two levels: the maximum level and another lower level or a level equal to 100 mW. This is the maximum power allowed in many countries outside North America, where power control is not mandatory. In North America the maximum power is 1W, and if utilized, power control would be mandatory. FHSS physical layer protocols and functions The FHSS physical layer consists of two protocol functions: One of them, the physical layer convergence function, provides the interface with the MAC and adapts the frame format of the MPDUs to the suitable format for FH transmission. The


87

second function, a physical mediumdependent function, defines the characteristics of transmission and reception of the data via radio. This separation is intended to simplify the interface between the MAC and the PHY, and to allow the MAC to operate with minimum dependencies on the transmission medium. Three functional entities are defined within the FHSS PHY. They are described as follows. ω

PMD sublayer: Provides the transmission interface between the transmitting and the receiving stations;

ω

PHY layer management entity (LME): Provides the management of the local PHY functions in conjunction with the MAC management entity;

ω

PLCP sublayer: Simplifies the interface of the MAC services with the PHY services.

The reference model for the physical layer is shown in Figure 3.15. Detailed descriptions of the PHY entities, protocols, and exchanged primitives are outside the scope of this text. The interested reader should refer to the standard [14]. The PHY functions can be grouped into two categories, described as follows. ω

Those that provide a frame structure for the transmission and reception of the MPDU and associated information. This is done by the PLCP.

ω

Those that transmit and receive the frames with the physical transmission characteristics defined in the previous section. The PMD

PLCP PHY LME PMD

Figure 3.15

Reference model for the PHY.

88


sublayer does this. The LME helps to synchronize the hopping sequences in all the stations within a given network. The MPDU length can be between 0 and 4,095 octets. This is a parameter needed both in transmission and reception. The required bit rate is a parameter needed in the transmission side, while the RSSI is needed in the reception side. The PLCP frame format is shown in Figure 3.16. The PLCP preamble and header are transmitted always at the basic speed of 1 Mbps. The PLCP_PDU can be transmitted at 1 Mbps or 2 Mbps. The fields are described as follows. ω

SYNC: This is an alternating zero-one sequence starting with zero and ending with one. It contains 80 bits and allows the receiver to detect a received signal, to synchronize the carrier and the clock, and to select an antenna if frequency diversity is employed.

ω

Start frame delimiter (SFD): This is a unique word that allows for the frame synchronism. The 16-bit pattern is: 0000 1100 1011 1101.

ω

PLCP-PDU length word (PLW): This indicates the number of octets contained in the MPDU packet. Valid states are from 0 to 4,095, with 12 bits.

ω

PLCP-PDU signaling field (PSF): Composed of four bits. The first is reserved for future applications. The three others indicate the bit rate: 000 is for 1 Mbps, and 010 is for 2 Mbps. The other six combinations are for the bit rates of 1.5, 2.5, 3, 3.5, 4, and 4.5 Mbps, but, at this moment, the physical layer specifications do not allow these speeds.

ω

Header error check (HEC): 16-bit field, obtained with a UITT CRC-16 generator polynomial.

Preamble SYNC 80 bits Figure 3.16

SFD 16 bits

Header PLW 12 bits

PSF 4 bits

PLCP frame format.

HEC 16 bits

PLCP - PDU

The IEEE 802.11 Standard ω

89

PLCP-PDU: Contains the data in the MPDU, scrambled and unbiased.

Data are scrambled using a synchronous scrambler of 127 bits in length. Its structure is shown in Figure 3.17. The same structure is used for descrambling in the receiver. Both scramblers are initialized at the beginning of the PLCP-PDU with an all-ones word. Scrambler data are grouped into 32-symbol packets. Stuff symbols (0 for 1 Mbps and 00 for 2 Mbps) are added at the beginning of each packet. To reduce the amount of bias (residual DC component due to imbalance between the number of zeroes and ones) a procedure is established to invert all the bits within one packet when bias is accumulating. Next, the physical layer for DS is presented. 3.3.6

IEEE 802.11 direct sequence physical layer

Radio transmission Transmission in the DSSS physical layer is defined in two modes: 1 Mbps and 2 Mbps. In both cases, the band between 2.4 and 2.5 GHz is structured in channels whose center frequency is shown in Table 3.7 for North America, Europe, and Japan. The symbol rate is always 1 Ms/s. The bit rate can be 1 Mbps, using BPSK modulation, or 2 Mbps with QPSK modulation. The chip rate is 11 Mchip/s. The spreading sequence is unique for all implementations, the Barker sequence being:

+1, −1, +1, +1, −1, +1, +1, +1, −1, −1, −1 Data In

T

T

T

T

T

T

T

(De) Scrambled Data Figure 3.17

Structure of the scrambler/descrambler.

90


Table 3.7 DSSS PHY Channels Channel

North America

Europe

Japan

1

2.412 GHz

2

2.417 GHz

3

2.422 GHz

2.422 GHz

4

2.427 GHz

2.427 GHz

5

2.432 GHz

2.432 GHz

6

2.437 GHz

2.437 GHz

7

2.442 GHz

2.442 GHz

8

2.447 GHz

2.447 GHz

9

2.452 GHz

2.452 GHz

10

2.457 GHz

2.457 GHz

11

2.462 GHz

2.462 GHz

12

2.484 GHz

The left-most chip shall be sent first in time and shall be aligned at the start of a transmitted symbol. As the spreading sequence period is equal to the symbol period, implementation is quite easy: For BPSK, the sequence is transmitted without changes when a 1 is sent, and the sequence is inverted to send a 0. The same can be said for QPSK, since each bit must modulate one of the two quadrature carriers. Synchronization is facilitated by the use of a very short code. Processing gain is equal to 11 or 10.4 dB. When several independent networks must operate in the same scenario, the processing gain states that the receiver can tolerate 10.4 dB more interference power than the threshold given by the modulation. Also, networks can share a scenario by using different channels separated by at least 30 MHz. More importantly, the wider bandwidth on the order of 22 MHz helps to improve the reception in the presence of multipath selective fading. A RAKE receiver would be able to distinguish signals arriving with time differences greater than 90 ns, to separate them, and to optimally combine them. No equalizers are needed. The modulation encoder for BPSK or QPSK is shown in Table 3.8. Transmit-to-receive switching time is fixed at 10 µs. Receive-totransmit switching time is 5 µs. Transmitter switching from on to off and


91

Table 3.8 Modulation Encoding Modulation

Symbol

Phase Change

DBPSK (1 Mbps)

1

0

0

π

00

0

01

π/2

11

π

10

−π/2

DQPSK (2 Mbps)

from off to on must be performed in less than 2 µs. Frequency tolerance is ±25 ppm for the channel frequency and for the chip rate. Transmitter power shall be at least 1 mW. Depending on regional regulations, the maximum transmitter power must be lower than 1W in the United States, lower than 100 mW (EIRP) in Europe, and lower than 10 mW/MHz in Japan. Power control is mandatory if the maximum power is higher than 100 mW. At least one of the power levels must be lower than or equal to 100 mW. The specified receiver sensitivity is −80 dBm for a BER of 8 × 10−2 with an MPDU length of 1,024 bytes. This is for 2-Mbps modulation. The maximum signal is −4 dBm. There are three modes of operation for the clear channel assessment (CCA). Receivers must be able to work with at least one of them: ω

Mode 1: Energy above threshold. The channel will be considered busy if energy above a given threshold is detected. The threshold is −80 dBm for transmission power greater than 100 mW, −76 dBm for transmission power between 50 and 100 mW, and −70 dBm for transmission power lower than 50 mW.

ω

Mode 2: Carrier sense only. The channel will be reported as busy if a valid DSSS signal is detected, with independence of its energy. If a PLCP header is detected, the channel will be considered busy until the end of the frame as indicated in the PLCP LENGTH field, even if reception is lost.

ω

Mode 3: Carrier sense with energy above threshold. This is a combination of both. The channel is considered busy if a valid DSSS signal is detected with energy greater than the threshold, calculated as in mode 1.

92


DSSS physical layer protocols and functions As in the FHSS PHY, the DSSS PHY consists of two protocol functions: the PLCP function and the PMD function. The first adapts the frame format of the MAC to a format suitable for transmission in this medium. The second performs the modulation and demodulation and associated functions. Three functional entities are defined within the DSSS PHY:

1. PMD sublayer: Provides the transmission interface between the transmitting and the receiving stations; 2. Physical LME: Provides the management of the local PHY functions in conjunction with the MAC management entity; 3. PLCP sublayer: Simplifies the interface of the MAC services with the PHY services. The PHY functions can be grouped into two categories: 1. Functions that provide a frame structure for the transmission and reception of the MPDUs and the associated information. This is done by the PLCP. 2. Transmission and reception of the frames with the physical transmission characteristics defined in the last paragraphs. This is done by the PMD. The LME helps the MAC in obtaining information on the state of the channel or the synchronization of the signals. The MPDU length can be between 0 and 4,095 octets. This parameter is exchanged with the MAC in transmission and in reception. The required bit rate is a parameter needed in the transmission side, while the RSSI is needed in the reception side. The PLCP frame format is shown in Figure 3.18. The PLCP preamble and header are always transmitted at the basic speed of 1 Mbps. The PLCP_PDU can be transmitted at 1 Mbps or 2 Mbps. The fields are described as follows. ω

SYNC: This is composed of a sequence of 128 bits equal to 1s. It is intended to provide all the synchronization needed.

ω

SFD: This is the 16-bit word F3A0H.


Preamble SYNC SFD 128 bits 16 bits

Figure 3.18

93

Header Signal Service Length CRC 8 bits 8 bits 16 bits 16 bits

MPDU

PLCP frame format.

ω

SIGNAL: This 8-bit field provides the receiver with the information on the modulation that will be used. The data rate is equal to the SIGNAL multiplied by 100 Kbps. Currently, only two possible values of data rate are specified. Thus, there are only two possible values of this field: 0AH for 1 Mbps and 14H for the 2 Mbps.

ω

SERVICE: This field is reserved for future use. The value 00H signifies an 802.11-device compliance.

ω

LENGTH: This field represents a 16-bit integer number that indicates the number of microseconds needed to transmit the MPDU.

ω

CRC: A UIT-R CRC 16 polynomial generator is used to provide error protection for the PLCP header.

ω

MPDU: This field contains all the data in the MPDU, in the same order.

Prior to transmission, all the bits in the PLCP frame are scrambled with a self-synchronized sequence with a length equal to 127 bits. The scrambler structure is shown in Figure 3.19, and the descrambler is shown in Figure 3.20. The scrambler is initialized to any word (except all zeroes) at the beginning of the frame. 3.3.7

Comparison of the FHSS and DSSS physical layers

FH and DS are two methods of spreading the spectrum in transmission systems. Both share some of the advantages of spread spectrum transmission, such as the lower probability of detection and interception, the higher resistance to interference, and the lower levels of interference over other systems. However, they are quite different in their modes of operation.

94


Scrambled data

T

T

T

T

T

T

T

Data in Figure 3.19

DSSS data scrambler.

Data in

T

T

T

T

T

T

T

Descrambled data Figure 3.20

DSSS data descrambler.

Many different arguments can be raised in favor of one or the other system. This section compares their applications in different scenarios to highlight their different behavior. First, lets review the different ways in which they combat channel perturbations arising from interference or from propagation impairments. FH relies on the assumption that perturbations will only affect some of the channels. When transmission is made in the perturbed channels, data will be received with poor or unacceptable quality. However, data transmitted in the unperturbed channels will be received without any difficulty. If the number of perturbed channels is small, by changing


95

the transmission frequency from time to time most of the data will be delivered without errors. In some systems the lost data can be recovered with the aid of interleaving and forward error coding. As interleaving is not considered in IEEE 802.11, the lost data must be retransmitted. Within a specified channel, the transmission is equivalent to narrowband and has no special protection against perturbations. DS, on the contrary, provides higher levels of protection within its transmission channels. First, it tolerates interference levels higher than the equivalent narrowband transmission in an amount equal to the processing gain. This is 10.4 dB in this standard. Second, it provides protection against multipath through the wider bandwidth and the possibility of the separation of signals arriving with delays higher than the chip period. Isolated networks can operate in DSSS or FHSS depending only on the propagation scenario. Small rooms can have multipath with a very small delay spread, and thus the channel coherence bandwidth can be larger than the DSSS bandwidth. In this case, FHSS might be preferable. Larger rooms, such as offices or industrial installations, are good scenarios for DSSS. Delayed paths can be separated in the receiver, and the extra protection against interference can help to provide constantly a very low BER even in the case of occasional interference from other systems. FHSS would suffer from the possibility of fading in some channels and the permanent need for retransmission. In additionbut not as importanttransmission time in DSSS is shorter, because there is no need to spend time changing the frequency of the channel. Networks deployed in scenarios where other equipment (802.11 or other types) operate in the same frequency band present the problem of controlling the amount of interference. If it can be guaranteed that in any receiver the interference level is below its threshold (with the processing gain taken into account), then DSSS is still a good alternative. However, interference levels higher than this threshold would make the receiver deaf to the desired signals, with no possibility of recovering. Thus, in complex scenarios without the possibility of controlling or even modeling the received interference power from other systems, FHSS must be the choice. Some bursts will be completely lost, but some of the data will be correctly received. To summarize, when DSSS works, it works well in a wide range of conditions. However, when it fails, as a result of propagation impairments affecting the full channel bandwidth, or because of high levels of interference, it fails dramatically, losing all the ability to receive signals. FHSS behavior is smoother, and more robust in some scenarios.

96


However, even in the best of conditions it presents the probability of losing some bursts, requiring retransmission.

3.4 Physical layer for IEEE 802.11 wireless LANS: Infrared systems IR radiation is a good alternative for sending information in closed areas. The IEEE-802.11 standard defines a PHY layer using this technology. Based on the common IEEE-802.11 MAC, the communication is based on a diffuse channel, although a LOS link can be used, with one or more receivers. This is the main difference with IrDA, which is only intended for point-to-point links. There are three main transmission modes for the emitter-to-receiver wireless optical link: LOS, one or several reflections on the walls and furniture, or by being received and resent by an active repeater. As the power reaching the receiver has a wide dynamic range, no value is set within the standard for the communication range, but in usual rooms or offices, about 10m is expected. Several people with their portable computers or PDS sitting around a common table, or in a not-too-large classroom, are the scenarios taken into account by the standard developers. Other scenarios are well-suited, toofor example, an airport hall or museum, airplane or train cabins, and bank offices. The network protocol is able to manage several nodes without the necessity of physical connections. If an external connection is needed, an AP should be implemented. Furthermore, a meeting room may have one or more active repeaters installed to improve the signal level anywhere inside it. This is the alternative to offering several RJ or coaxial sockets for the participants. The reduced range can become advantageous if the data should be confined to the room or if confidentiality is a goal. This is the case in bank offices or meeting rooms. Due to its intrinsic short range, no regulation has been made on IR frequency allocation for transmission. On the other hand, several eyesafety regulations limit the power emission of IR devices. 3.4.1

Description

The PHY is divided into two sublayers: PLCP and PMD. The PLCP interfaces with MAC to convert MAC packets into the format used by the circuitry of the PMD. The PMD sublayer performs no data processing, so if


97

new hardware techniques or devices are developed, they can be easily implemented. Both sublayers can be implemented together without a physical frontier between them. For example, one chip can set up the synchronization frame, CRC checksum, PPM encoding, and energy detection. Sections 3.4.2 and 3.4.3 discuss the main characteristics of both sublayers. 3.4.2

The physical layer convergence procedure (IR-PLCP)

This sublayer is where MPDUs are converted to electrical signals to be applied to PMD devices. It has two purposes: to assure an error-free transmission, and to simplify the reception procedure. Inserting, before the MPDU, a preamble and a header makes this. The preamble consists of a long series of pulses to synchronize the receiver clock, to achieve good data extraction. On the other hand, the header includes all the information needed on the receiver PHY. The basic optical signal is a 250-ns pulse. This period is known as slot time. Although two data rates are defined, the pulse length is the same so the optical receiver can be optimized at this pulse duration. Most of the information is coded in PPM format, but other signals, mainly in the PLCP preamble and header cannot be modulated. In any case, the pulse duration is always the same. For example, synchronism pulses are a clock signal of 2 MHz (i.e., a train of many 250-ns pulses) used to synchronize the receiver clock; if the signal were PPM-codified, it could not be decoded because the receiver is out of synchronism at this time. The frame sent to PMD, or received from it, has three parts: the PLCP preamble, the PLCP header, and the PSDU. This frame is named the PLCPDU. This includes some pulses to synchronize the receiver (SYNC) and the SFD. The SYNC part is a clock square signal of 2 MHz that lasts between 57 and 73 slots. The last slot has to be empty (i.e., have no pulse). Its main goal is to allow the receiver clock to synchronize, although it can also be used to estimate the signal-to-noise ratio and automatic gain control, or to choose the receiver if diversity is implemented. It is followed by the SFD, which is always the nibble 1001. This signal marks the beginning of the frame and allows the symbol synchronization. These signals are made of single pulses; they are not modulated in PPM format. In fact, there is no PPM symbol with these sequences of pulses. PLCP preamble

98


PLCP header The header gives the information needed to translate the PPM symbols into bytes. It includes the data rate (DR), the DC level adjustment (DCLA), the length of the data in PSDU (LENGTH), and a checksum of the LENGTH parameter (CRC). The DR block is three slots long and can take the values listed in Table 3.9. The DR block values are not modulated in PPM. After DR there is a sequence of pulses to allow the receiver to adjust the DC level of the remaining signals. This is called DC level adjustment (DCLA), and it is needed because the mean value of 4PPM and 16PPM is not the same. The DC level of these signals would be the same as an average PPM signal at the data rate used. Their values are listed in Table 3.10, and only these two values are permitted. The third block is LENGTH, that is the number of bytes of the MPDU. It is a 16-bit unsigned integer and is the first field to be PPM-modulated. The LSB is transmitted first. To assure that the previous value is correctly received a checksum of 16 bits is sent. This is calculated using the polynomial x16 + x12 + x5 + 1. This field is also PPM-modulated.

This is the data coming from the MAC. Its length is defined by the LENGTH field, can vary between 0 and 2,500 bytes, and is PPM-modulated. The LSB is sent first. The process of transforming bits to PPM symbols is described in Section 3.4.3.

PSDU

Table 3.9 DR Block Values Data Rate

Value

1 Mbps

000

2 Mbps

001

Table 3.10 DCLA Values for Both 1 and 2 Mbps Data Rates Data Rate

Value

1 Mbps

0000 0000 1000 0000 0000 0000 1000 0000

2 Mbps

0010 0010 0010 0010 0010 0010 0010 0010


3.4.3

99

The IR physical medium sublayer (IR-PMD)

This layer only describes the signals format and the minimum specifications needed for communication between two IEEE 802.11-IRconformant devices. Two data rates are defined: 1 and 2 Mbps. Both data rates are implemented for the detectors, but 2 Mbps is optional for emitters. As the emitter power consumption depends heavily on the data rate, the choice is based on battery life for portable devices. The most common implementation includes a speed and rate that are selected depending on the battery charge. Also, at the lower speed the receiver sensitivity is larger so that one conversation can be started at one speed, and if the channel characteristics change, the other one can be used. Characteristics of the signal used The electrical and optical signals are based on a 250-ns pulse. This defines the slot time both for modulated and unmodulated signals. By using just one fixed-length pulse data, the receiver and emitter can be optimized for this signal, independent of the data rate. Two PPM schemes (see the next section) are used for sending the data, one for each rate. Nevertheless, other signals have to be sent unmodulated, at the most obvious data rate. (The data cannot be extracted until the PPM format is established.) A good synchronization is required for PPM data recovering, so a long clock signal is sent at the beginning of each frame. Pulse position modulation (PPM) This is the best modulation method for low- and medium-speed optical signals. In this method, a large pulse is sent. Its amplitude and shape are not important. The important factor is its delay relative to a symbol clock. The maximum current that can be applied to an LED has two values: maximum DC current and maximum pulsed current. The first value is due to the thermal dissipation capacity of the device: the larger the current, the larger the heat generated and the device temperature. On the other hand, short and long current pulses can damage the device because of current channeling, and other processes. In any case, the maximum pulsed current is usually 5 to 10 times larger than DC maximum current. Of course, the LED should be switched off for a period long enough to lose the heat generated while in the on state. PPM offers a duty cycle low enough to surpass the DC maximum current. Using, for example, a 16 PPM, the mean current will be about one-sixteenth of the peak current. So a 100 mA LED can be driven with 1A pulses, without degrading the device. On the receivers photodiode, the signal will be almost 10

100


times larger. Unfortunately, the bandwidth needed will be larger too, so the benefits are not so important on the signal-to-noise ratio. There is another factor to be taken into account: the electrical power needed to turn on the LED. Usually wireless communications are intended to be used on battery-powered equipment. If large currents are switched on and off in short pulses, the battery life and the charge duration will be reduced, and the electromagnetic interference (EMI) propagated through power supply connections could be a problem. To maintain an almost constant discharge regime, the LED driver charges a large capacitor while in the off state; this capacitor supplies the large current pulses while in the on state. With this technique, the pulse shape cannot be controlled too tightly, but this is not important in a PPM scheme. Figure 3.21 presents the electrical spectrum of a PPM signal. In Figure 3.21, the electrical bandwidth is much larger than the data rate, and this modulation cannot be used if the bandwidth is a scarce resource, as in RF. Fortunately, the optical spectrum is unregulated because the optical signals are blocked by walls, so a large amount of interference is impossible. As more and more IR equipment is used, some interference will be unavoidable, so three options are feasible for the forthcoming developments. The first one, now used by IrDA, is to establish the communication based on LOS links; the second one is to use another wavelength range; and the third is to use a carrier modulation over 20 MHz. The second option is handicapped by the high cost of optoelectronic devices for longer wavelengths; as the short history of electronic technology teaches us, the prices will fall as the number of devices increases. Accordingly, this problem will diminish if the new applications are successful. Going to a larger wavelength has a second advantage: The optical power level can be increased because the sensitivity and transparency of the human eye is lower. As a result, they are safer than shorter wavelengths. The carrier-modulated option will take advantage of all the circuitry developed for RF, but it will need faster, and again more expensive, emitters and detectors. In conclusion, IEEE 802.11-IR has taken the most convenient position. The problems will exist for future IR standards. With this in mind, PPM is the best option. PPM modulation is described as follows. Let T0 be the bit period (i.e., the inverse of the data rate). The time needed to send k bits will be k · T0 seconds. If we divide this time in 2k slices (slots), every one will represent the value of every k-bit combination. The duration of one slot is Ts = k · T0 / 2k seconds, so no more than a few bits can be grouped before the slot time becomes too short.


101

Pxx - X power spectral density

102

4-PPM 10

0

16-PPM

10-2

10-4

10-6

10-8

0

0.1

Figure 3.21

0.2

0.3

0.4 0.5 0.6 Frequency

0.7

0.8

0.9

1

Electrical spectrum of a PPM signal.

A single pulse placed into a time slot represents the value of the k bits. Because of this, synchronization is very important: An error in the delay measurement will produce a different bit combination. IEEE 802.11-IR uses 4-PPM and 16-PPM modulation for 2- and 1-Mbps transmission. The number indicates the number of slots. So one symbol (pulse) defines two bits at 2-Mbps, and four bits at 1-Mbps. With the synchronization being so critical, the pulse position is not just the plain bit group value. Instead, a Gray coding scheme is used so that one slot timing error forces one bit error. With these figures, it is easily seen that Ts is 250 ns in both cases. What is different is the duty cycle: one-sixteenth at 1-Mbps, and a quarter at 2-Mbps. If a given amount of data is to be sent, the energy requirement and the time needed for transmission are different. Table 3.11 presents these requirements, relative to 1 Mbps. The basic data unit is the byte (octet), but the symbols are composed of two or four bits, so bytes have to be broken up before their transmission. The rule defined is: Send blocks of bits starting with the less significant ones, but keeping the internal order inside the block.

102


Table 3.11 Energy and Time Requirements Data Rate

Energy

Time

1 Mbps

1

1

2 Mbps

4

0.5

Figure 3.22 shows the order for 2-Mbps data rate. Notice that the pair order is kept, so the second symbol transmitted will be the one corresponding to the values of the bit pair 32. The same rule is applied to 1 Mbps, but two 4-bit blocks are set up for each byte. Optical transmitters Regarding the optical parameters, two kinds of devices are described: The first is mainly oriented toward active repeaters on the ceiling, while the second is to be implemented in the computers, perhaps as a PCMCIA card. They differ only in the emission properties, having common receiver characteristics. The first one has a high power level (2W); this value will exhaust a portable computer battery in a few minutes. Nevertheless, it is not intended for mobile equipment but rather for being placed on the ceiling to act as an active repeater (AP), so it can be powered from the mains by a power supply. The beam pattern has a torus-like shape. A spherical or Lambertian profile would send too much power to the floor lying just under the emitter and much less to the walls. Because the most probable placement of receivers is on a table at a fixed height and several meters apart from the

Byte 76543210

Figure 3.22

Order of transmission: 10 32 54 76

lsb

First

msb

Last

Bit ordering in a 2-Mbps transmission.


103

vertical of the emitter (see Figure 3.23), the beam has been designed to cover these positions without wasting power in other directions. Nevertheless, as the emitted power is high, reflections on walls and furniture can reach the receivers with enough power to be detected. The second profile (or mask as it is defined in the standard) emits less peak power (500 mW), and the beam is closer to a cone of 60 degrees. This beam can be aimed at the repeater or at other devices; in fact, it can be oriented to any white surface that serves as a passive reflector or diffuser. This one is intended to be used in a portable computer, PDS, or any handheld device. Although the power is not so low, it should be kept in mind that only when the device is emitting is high power drawn. In the common use of this equipment, traffic is heavily asymmetric (i.e., portable devices are to receive most of the information), and only several ACKs are returned. The option to reduce the data rate to 1 Mbps (with a lower duty cycle) while sending could be a good opportunity for batterypowered devices, while receiving at 2 Mbps. The wavelength used is in the 850950-nm range, the one used by IrDA and TV remote controllers, so LED devices are very cheap. To avoid interference with other (future) devices operating at higher speed, the electrical power spectrum of the pulses should fall by 20 dB at 15 MHz. Optical receivers Two power-related parameters are defined for the receiver: sensitivity and dynamical range. The sensitivity is defined as the minimum power density on the receiver surface to achieve a FER of 4 · 10−5

Longer distance Large angle 2

Figure 3.23

Short distance Small angle 1

The optical intensity for direction 2 has to be larger than for 1.

104


in a 512-byte frame. This optical power is disturbed by an unmodulated background IR source, so the receiver should be able to recover the signal while other sources, mainly the IR spectrum of illumination lamps, are present. The sensitivity levels are 2 · 10−5 mW/cm2 at 1 Mbps and 8 · 10−5 mW/cm2 at 2 Mbps. These values are minima and can be improved to achieve longer distances. The optical power on the receiver can take a wide range of values and depends on the relative orientation between emitter and receiver, furniture, walls, and the people using the system. Based on this, a large (30-dB) dynamical range is mandatory for the receiver circuit. That fact is also responsible for no range distance having been defined in the standard. The PHY uses three signals to describe the average availability. These signals are energy detect (ED), carrier sense, and CCA. ED is triggered when the optical energy changes more than 1 W/cm2 in the 110-MHz band. Carrier sense is asserted when a preamble signal has locked the receiver clock. The receiver clock circuitry may be able to lock it, although the level of the signal is not enough to trigger ED; in this case carrier sense will be asserted, and ED will not. The CCA signal has two states: IDLE if the channel is free and BUSY if the opposite is true. When both ED and carrier sense are not asserted, CCA will assume the IDLE state. Nevertheless, other systems (remote controls, electrically switched fluorescent lamps, etc.) can maintain ED asserted although the channel is free. In this case, after a period of time with ED asserted but without carry detection, the CCA will change to IDLE. The state of CCA is sent to MAC, so it can manage all the possible events.

3.5

Conclusions and applications

The standard IEEE 802.11 is the first serious and universally accepted standard for WLAN. It covers the high-quality area of mobile data communications. As a living standard, several improvements have been developed. The 802.11a and 802.11b versions take advantage of new RF channels to increase the data rate to 20 or 25 Mbps in the 5.2- and 17.1-GHz bands. Because the main objectives of WLAN are PDAs and portable computers, a great number of these products are implemented in PCMCIA


105

cards. As the features of these operating systems promote networking, many services such as file transfer, e-mail, and, of course, Internet connection are available to users. Other equipment can also take advantage of mobile networking. For example, moving robots can be controlled from a central computer in a more flexible way. Other scenarios include meeting rooms, quality control in industries, academic rooms, hospitals, and libraries. The home scenario is not well suited for 802.11 networks because of the price. Simpler and cheaper solutions in which all the appliances are controlled by a central unit, are under development for home use. The possibility of e-mailing and Web browsing offered by new mobile telephones is not, in any sense, an alternative to wireless networking. IEEE 802.11 networks are intended to be implemented as a facility by companies for their workers. It should never be forgotten that they are based on a LAN standard. On the other hand, smart mobile phones have improved information processing capabilities. They do not allow users to modify or resend documents of several pages, including figures, graphs, and tables. The main makers of WLAN products have set up the WLAN Association (WLANA) to extend and publicize the capabilities of these networks.

References [1]

Pahlavan, K., and A. Levesque, Wireless Information Networks, New York: J. Wiley & Sons, 1995.

[2]

Barry, J., Wireless Infrared Communications, Dordrecht, Netherlands: Kluwer Academic Publishing, 1994.

[3]

Kahn, J., Wireless Infrared Communications, Tutorial at PIMRC 96, Taipei, Taiwan, ROC.

[4]

Kahn, J. M., and J. R. Barry, Wireless Infrared Communications, Proceedings of the IEEE, Vol. 85, No. 2, Feb. 1997, pp. 265298.

[5]

Pérez-Jiménez, R., V. M. Melián, and M. J. Betancor, Analysis of Multipath Impulse Response of Diffuse and Quasi-Diffuse Optical Links for IR-WLAN, Proceedings IEEE INFOCOM95, April 46, 1995, Boston, MA, pp. 7d.3.17d.3.7.

[6]

Chen, K.C., Medium Access Control of Wireless LANs for Mobile Computing, IEEE Network, Sept.Oct. 1994, pp. 5063.

106


[7] Crow, B. P., et al., IEEE 802.11 Wireless Local Area Networks, IEEE Comm., Sept. 1997, pp. 116126. [8] LaMaire, R., et al., Spread Spectrum Access Methods for Wireless Communications, IEEE Comm., Jan. 1995, pp. 5867. [9] Bantz, D., and F. Bauchot, Wireless LAN Design Alternatives, IEEE Network, March/Apr. 1994, pp. 4353. [10] Kamerman, A. D., Spread Spectrum Schemes for Microwave Frequency WLANs, Microwave Journal, Feb. 1997, pp. 8090. [11] Abramson, N., Multiple Access in Wireless Digital Networks, Proceedings IEEE, Vol. 82, No. 9, Sept. 1994, pp. 13601370. [12] Kohno, R., R. Meidan, and B. Milstein, Spread Spectrum Access Methods for Wireless Communications, IEEE Comm., Jan. 1995, pp. 5867. [13] Elmirghani, J. M. H., and R. A. Cryan, Indoor Infrared Wireless Networks Utilising PPM CDMA, Proceedings ICC 94, Vol. 2, June 1995, pp. 731734. [14] IEEE 802.11 Standard, New Jersey: IEEE Press, 1999. [15] Wickelgrem, I. J., Local Area Networks Go Wireless, IEEE Spectrum, Sept. 1996, pp. 3440. [16] A. Santamaría, and F. J. López-Hernández, Wireless LAN Systems, Norwood, MA: Artech House, 1994. [17] Diepstraten, W., G. Ennis, and P. Belanger, Distributed Foundation Wireless MAC, IEEE P802.11-93/190. [18] Valadas, R. T., J. M. Brazio, and A. M. Oliveira, Throughput Performance of Nonpersistent CSMA/CD Quasidiffuse Infrared Local Area Network Under an Imperfect Average Power Sensing Collision Detection Method, Proceedings ICC 93, Vol. 1, May 1993, pp. 567572. [19] Jacquet, P., and P. Muhlethaler, CSMA-Radio Mobile MAC Proposal, IEEE P802.11/93-99. [20] Tobagi, F., and L. Kleinrock, A Packet Switching in Radio Channels, Part I & II, IEEE Trans. on Communications, 1975. [21] Ahmadi, H., A. Krishna, and R. O. LaMaire, Design Issues in Wireless LANs, Journal on High-Speed Networks, Vol. 5, No. 1, 1996, pp. 87104. [22] LaMaire, R., A. Krishna, and H. Ahmadi, Analysis of a Wireless MAC Protocol with Client-Server Traffic, Proc. IEEE INFOCOM, 1993, pp. 429438. [23] Chaaya, H. S., and S. Gupta, Performance of Asynchronous Data Transfer Methods of IEEE 802.11 MAC Protocol, IEEE Personal Communications, Oct. 1996, pp. 815.


107

[24] Chaaya, H. S., and S. Gupta, Throughput and Fairness Properties of Asynchronous Data Transfer Methods of IEEE 802.11 MAC Protocol, Proceedings of the IEEE PIMRC 95, 1995, pp. 613617. [25] Minnet, P., HIPERLAN: A European Standard for 5 GHz and 17 GHz Wireless LAN, Microwave Engineering Europe, Oct. 1995, pp. 6166. [26] ETSI TC-RES Radio Equipment and System (RES) HIPERLAN: Functional Specifications, ETSI 06921, Sofia, Antipolis, France 1995. [27] Wilkinson, T. G. C., T. Phips, and S. K. Barton, A Report on HIPERLAN Standardisation, International Journal on Wireless Information Networks, Vol. 2, Apr. 1995, pp. 99120. [28] HIPERLAN MAC Presentation, IEEE 802 Plenary Meeting, Montreal, Canada, Nov. 1997 [29] Simon, M. K., et al., Spread Spectrum Communications Handbook, New York: McGraw-Hill, 1994. [30] Pickholtz, R. L., D. L. Schilling, and L. B. Milstein, Theory of Spread-Spectrum Communications: A Tutorial, IEEE Trans. on Communications, Vol. COM-30, No. 2, May 1982, pp. 855884. [31] Holmes, J. K., Coherent Spread Spectrum Communications, Wiley, New York: 1982. [32] Viterbi, A. J., CDMA Principles of Spread Spectrum Communications, Reading, MA: Addison-Wesley, 1995. [33] Gibson, J. D. (ed.), The Mobile Communications Handbook, Boca Raton, FL: CRC Press Inc., 1996. [34] Shannon, C. E., Communications in the Presence of Noise, Proceedings of IRE, Vol. 37, Jan. 1949, pp. 10-21 [35] de Gaudenzi, R., C. Elia, and R. Viola, Band Limited Quasi-Synchronous CDMA: A Novel Satellite Access Technique for Mobile and Personal Communications Systems, IEEE Journal on Selected Areas in Communications, Vol. 10, No. 2, Feb. 1992, pp. 468481.

.

CHAPTER

4

Contents 4.1 Introduction: Terminology 4.2

Physical layer (PHY)

4.3 HIPERLAN channel access control (CAC) 4.4 HIPERLAN medium access control (MAC) 4.5 Conclusions on HIPERLAN type 1 4.6 Future BRAN standards

The HIPERLAN Standard J. M. Riera

4.1

Introduction: Terminology

In 1992, the Conference Européene des Administration des Postes et des Télécommunications (CEPT) allocated the frequency bands in the 5.15 GHz to 5.30-GHz and 17.1 GHz to 17.3-GHz bands [1] for the deployment of high-speed radio LANs. The use of these bands is permitted without frequency planning and/or individual licensing. Users may not claim protection and must not interfere with other users. Equipment must comply with the technical characteristics defined by the ETSI. Part of the 5-GHz band, from 5.15 GHz to 5.25 GHz, is available on a pan-European basis. The remaining 50 MHz, from 5.25 GHz to 5.30 GHz, is available on a national basis. As a way of providing economies of scale that would launch the development of a market for these products, the ETSI has developed a standard known as HIPERLAN type 1 [2], which operates in the 5-GHz

109

110


band with bit rates of 23 Mbps. This standard is intended to allow the establishment of two types of HIPERLANs: ω

Predefined LANs that can be made to form part of, or supplement, fixed LANs;

ω

Ad hoc HIPERLANs that can be rapidly configured for ad hoc functionality.

In addition, efforts have been made in the definition of HIPERLAN type 1 to do the following: ω

Provide a PHY capable of working at bit rates comparable to or even higher than fixed LANs;

ω

Provide a multiple access scheme that ensures the quick delivery of time-bounded traffic, while maintaining good spectrum efficiency and good characteristics for asymmetric traffic;

ω

Provide a means of differentiating HIPERLANs that may overlap and use the same frequency band without mingling traffic;

ω

Provide greater HIPERLAN coverage than the radio range of individual nodes with a means of resolving the potential problem of hidden nodes.

A HIPERLAN is a high performance radio local area network in which all nodes communicate using a single shared communication channel. It has the following properties [2]: ω

It provides a service that is compatible with the ISO MAC service definition;

ω

Its operations are compatible with the ISO MAC bridges specifications for interconnections with other LANs;

ω

It may be deployed in a prearranged or an ad hoc fashion;

ω

It supports node mobility;

ω

It may have coverage beyond the radio range limitation of a single node;

ω

It supports both asynchronous and time-bounded communications by means of a channel access mechanism with priorities providing hierarchical independence of performance.

The HIPERLAN Standard ω

111

Its nodes may attempt to conserve power in communications by arranging themselves when they need to be active for reception.

The main features of HIPERLAN can be summarized as follows: ω

Routing procedures that allow the dynamic establishment of routes that link nodes that are not within each others radio range, so that they may exchange communications. With these procedures, the HIPERLAN range is virtually unlimited.

ω

Power-saving functions that allow terminals to have some of their circuits switched off most of the time but to be switched on for small duty cycles.

ω

Elimination-yield nonpreemptive priority multiple access (EYNPMA), an advanced channel access protocol that assigns priorities to traffic on the basis of its specification regarding delivery time, thus reducing the probability of collisions to a minimum by letting nodes contend for the use of the channel before sending the communication. The contention is based on measurements on the state of the channel (busy/idle) with an adaptive receiving threshold. Provision is made for detection of, and not interference with, hidden nodes.

ω

A modulation scheme with two different bit rates: The low bit rate (LBR) of roughly 1.5 Mbps is used for service information, and indoor channels can be received without equalization. The high bit rate (HBR) of 23 Mbps allows for the rapid transmission of data. The receivers need equalizers for the HBR. The modulation format is FSK for the LBR and GMSK for the HBR.

The HIPERLAN 1 standard defines three layers: the MAC sublayer, the CAC sublayer, and the PHY. The three correspond to the PHY and DLLs of the OSI reference model. HIPERLAN applications are considered as protocols belonging to a higher layer. Figure 4.1 shows the HIPERLAN reference model and its relation with the OSI reference model. When data transmission is needed, applications (users) access HIPERLAN at the MAC service AP (MSAP) and demand the transmission of their MSDUs. The MAC service is based on HIPERLAN MAC PDUs (HMPDUs). Some HMPDUs carry user data, from the MSDUs. Others are used to exchange information between the MAC entities of different terminals. MAC entities have to maintain a number of databases with information needed for the routing functions. This information is

112


Applications

Higher layers

Medium access control (MAC) sublayer

Channel access control (CAC) sublayer

Data link and Physical layers

Physical (PHY) layer

HIPERLAN Figure. 4.1

OSI

HIPERLAN reference model.

dynamically updated on the basis of received data from other MAC entities. HMPDUs are delivered to the HIPERLAN CAC service AP (HCSAP) for the CAC service. This service is based on the HIPERLAN CAC PDUs (HCPDUs). Some of them include data proceeding from the higher layers (user data or MAC data) and others are generated by the CAC entity as part of the implementation of the channel access protocol. HCPDUs are delivered to the PHY, which codes them as bursts and then transfers them to other HIPERLAN nodes via radio. All users within a single HIPERLAN use the same radio channel, with sharing based on time division multiple access. The distribution of time between the users is part of the CAC service. On the receiving side, the reverse process takes place. The PHY receives the radio bursts and decodes them, and if their destination is in the same terminal, the HCPDU delivers them to the CAC. If the HCPDU contains data for the higher layers, the HMPDU delivers them to the


113

MSDU MSAP

HMPDU

MSDU MAC service

MAC protocol

HCSDU HCSAP

HCPDU

Data burst Figure 4.2

MSAP

HMPDU HCSDU

CAC service

CAC protocol

PHY protocol

HCSAP

HCPDU

Data burst

HIPERLAN communication model.

MAC, and, eventually, the MSDU is delivered to the user. Figure 4.2 shows the reference communication model. The following sections describe the HIPERLAN PHY (Section 4.2), CAC (Section 4.3), and MAC (Section 4.4), paying special attention to those features specific to this standard.

4.2

Physical layer (PHY)

4.2.1

Introduction

The main goals of the HIPERLAN PHY can be summarized as follows: ω

To establish a physical link to deliver data from a transmitter to one or several receivers using the modulation formats described for the LBR and HBR transmission and the techniques for error correction as described in the standard;

ω

To assist the multiple access scheme by measuring the channel status according to predefined rules and maintaining an adaptive threshold to determine whether the channel is busy.

114


The standard [2] defines the following tasks for the PHY: ω

To modulate and demodulate radio carriers with a bit stream of a defined instantaneous rate to create an RF link;

ω

To acquire and maintain bit and burst synchronization between transmitters (TXs) and receivers (RXs);

ω

To transmit or receive a defined number of bits at a requested time and on a particular carrier frequency;

ω

To add and remove the synchronization sequence;

ω

To encode and decode the FEC scheme;

ω

To measure the received signal strength;

ω

To decide whether a channel is idle or busy for the purposes of deferral during channel access attempts;

ω

To maintain a defer threshold.

The first five tasks are related to data transmission. The last three are related to the multiple access scheme. 4.2.2

Transmission characteristics

There are five channels in the frequency band from 5,150 to 5,300 MHz. Table 4.1 lists the carrier frequencies nominal values. Channels 0, 1, and 2 are defined as default channels and are available in all countries under the CEPT regulations. The availability of channels 3 and 4 depends on national regulations. As HIPERLAN terminals can be taken to different countries, there are means of informing the terminals on the availability of these two channels. However, the whole problem of Table 4.1 HIPERLAN 1 Carrier Frequencies Carrier Number

Frequency (MHz)

0

5176.4680

1

5199.9974

2

5223.5268

3

5247.0562

4

5270.5856


115

channel selection for a network is not addressed in this standard. It is left to the higher network layers. All terminals belonging to an individual HIPERLAN must transmit and receive on the same channel. Transmitter frequency accuracy must be better than 10 ppm with respect to the nominal carrier frequency. There are three types of transmitters: Class A transmitters always have a transmission power of +10 dBm. Class B transmitters have a maximum power of +20 dBm but have implemented power control with two levels: +10 dBm and +20 dBm. Class C transmitters have maximum power of +30 dBm and power control with three levels, +10 dBm, +20 dBm, and +30 dBm. The power must be within 5 dB of the specified level but never exceed 1W. All values are in effective isotropic radiated peak envelope power (EIRPEP). As the transmission is in bursts, the rise and fall times are specified. The burst is shown in Figure 4.3. Outside a burst period, transmitted power must be at least 60 dB under the nominal power. Rise time to the nominal power must be lower than 2.5 µs, but only in the last 1.5 µs may the power exceed the nominal power minus 20 dB. Fall time must be lower than 4 µs, but the power must go below the nominal power minus 20 dB in 3 µs after the end of the burst. During the burst, power variations must be smaller than ±2 dB.

Power +2 dB Mean power -2 dB

-20 dB

-60 dB

Start of the burst 1.5 µs 2.5 µs

Figure 4.3

Burst characteristics.

End of the burst 3 µs 4 µs

Time

116


LBR modulation is two levels of frequency shift keying (FSK). The signaling frequencies are equal to the center frequency plus (for a one) or minus (for a zero) 368 KHz. The nominal bit rate is 1.470588 Mbps. No filtering is performed on the baseband signal. HBR modulation is Gaussian minimum shift keying (GMSK). This is a variety of FSK in which the frequency deviation from the carrier is set at 1/4Th , Th being the bit period for the HBR. The baseband data are filtered with a Gaussian filter. The Gaussian filter bandwidth is usually related to the bit period. The parameter of the GMSK modulation is the product of both BTs. In HIPERLAN BT = 0.3. The nominal bit rate is 23.5294 Mbps. This is 16 times the LBR. The advantages of GMSK are the power efficiency (because the signal has constant envelope and nonlinear amplifiers can be used) and the low level of adjacent channel power. The disadvantages are the lower immunity from noise (due to the baseband filtering, the maximum frequency deviation is only reached after a certain number of consecutive ones or zeroes) and some degree of intersymbol interference. Delay spread in indoor channels is usually smaller than the bit period in the LBR format but higher than the bit period in the HBR. For this reason, HIPERLAN receivers need equalizers that cancel the intersymbol interference due to the channel and the modulation. Though the equalizers are not described in the standard, a synchronization sequence is transmitted in the HBR part of the bursts that can be used as a training sequence for the equalizers [3]. The maximum length of the HBR part of the bursts has been calculated so that the channel transfer function does not change significantly in the interval. 4.2.3

Data bursts

Two types of data burst are defined in the standard. The LBR data burst, which encodes an LBR HIPERLAN CAC PDU (HCPDU), and the LBR-HBR data burst, which encodes an LBR-HBR HCPDU. Thus, the HBR transmission is always preceded by an LBR preamble. The bits of an LBR HPCDU modulate the carrier in FSK to form an LBR burst. This is only used for the Acknowledgement HCPDU (AK-HCPDU), whose structure will be explained in the CAC section. The LBR-HBR data burst has two parts. The LBR part is made up of a number NL of bitsthe ones in the LBR part of the LBR-HBR HCPDU that modulate in FSK the carrier. The HBR part is made up of a number NH of bits, which modulate the carrier in GMSK. The NH bits encode the information in the HBR part of the LBR-HBR HCPDU but also include


117

error correction features and a synchronization sequence. Figure 4.4 shows the structure of the LBR-HBR burst. The HBR stream includes the following: ω

A synchronization and training sequence of 450 bits. This sequence is fixed for all transmitted HIPERLANs.

ω

A number, m, of data blocks of 496 bits. The minimum value of m is 1, and the maximum is 47. This maximum is determined for speeds of up to 1.4 m/s. For higher speeds this maximum value should be reduced. This m × 496 bits encodes the HBR part of the LBR-HBR HCPDU that consists of m × 416 bits. The operations of forward error encoding, interleaving, bit toggling, and differential precoding are carried out on the data.

The maximum duration of the HBR part is 1 ms. At 1.4 m/s, the displacement is 1.4 mm or approximately λ/40. The operations on the data blocks are described as follows: ω

The data block of 416 bits is segmented into 16 segments of 26 bits.

ω

Each 26-bit segment is encoded with a (31,26) Bose-ChaudhuriHocquenghem (BCH) code. This means that five redundancy bits are added to the 26 data bits. This code is able to correct two random errors in a word.

ω

The data block is made up of 16 words of 31 bits, with a total of 496 bits, but interleaved. The process is similar to writing the words in a 16 × 31 matrix, so that each word is a row in the matrix, and the matrix is read by columns. The first 16 bits of the data block are the 16 first bits of the 16 words. The next 16 bits of the date block are the 16 second bits of the 16 words, and so on.

LBR part NL bits

Synchr. & training sequence 450 bits

Data block 0 Data block 1 496 bits 496 bits HBR part

Figure 4.4

LBR-HBR burst structure.

Data block m-1 496 bits

118


This combination of coding and interleaving should be able to correct burst errors with a length of up to 32 bits, provided there are no more errors in the data block.

ω

Then the m resulting data blocks are joined. Bits with odd order numbers are inverted, and differential precoding is done.

All these operations are carried out on the data. The synchronization and training sequence and the LBR part are added at the end of the process. 4.2.4

Channel access bursts

These bursts are used in the first phases of the EY-NPMA, when nodes must contend before transmitting their traffic. The bursts are formed by the repetition, using the HBR, of the following 32-bit sequence: 11111010100010011100000110010110 Section 4.2.5 explains the duration of the channel access bursts, which is determined by the CAC protocol. 4.2.5

Receiver characteristics

Receiver sensitivity is defined in HIPERLAN for the different bit rates as follows: ω

For LBR data bursts, it is the power level at the receiver RF input at which the proportion of received LBR HCPDUs with error is 1%.

ω

For the LBR part of LBR-HBR data bursts, it is the power level at the receiver RF input at which the proportion of received LBR parts of the LBR-HBR data bursts with error is 1%.

ω

For the HBR part of LBR-HBR data bursts, it is the power level at the receiver RF input at which the proportion of received HBR parts with uncorrected errors is 1%. This is measured with HBR parts of 10 blocks.

Thus, there are three different sensitivity values. According to the highest of the three, receivers are classified as follows: Class A receivers have −50 dBm; class B receivers have −60 dBm; and class C receivers have −70 dBm. These figures are for isotropic reception antennas.


119

The maximum input level for operation is −20 dBm, and the maximum input level without damage is 0 dBm. The maximum number of consecutive ones or zeroes without impairment of performance is 128. 4.2.6

Compatibility between transmitter and receiver classes

The three types of transmitter and three types of receivers can help manufacturers offer different degrees of quality. Some pieces of equipment may be cheaper than others or use less power, thus making them more appropriate for portable equipment. However, not all combinations are allowed. Higher-power transmitters must always be associated to higher-sensitivity receivers. Thus, a HIPERLAN terminal with a type C transmitter (highest power) must include a type C receiver (highest sensitivity). A type B transmitter may be associated to a type B or a type C receiver, and a type C transmitter (lowest power) can be associated to any type of receiver. 4.2.7

Establishing a defer threshold

To assist the multiple access protocol, the HIPERLAN receivers are asked to determine at times the state of the channel and classify it as busy or idle. The receivers must be able to do the following: ω

Measure the received signal level.

ω

Maintain a defer threshold so that when the received signal level is lower than this threshold the channel is detected as idle and is otherwise assumed to be busy.

The received signal level is measured in the bandwidth used for the HBR and expressed in terms of signal level number (SLN), ranging from 0 to 31. SLN = 1 corresponds to −75 dBm. SLN = 31 corresponds to −25 dBm. On average, the unit SLN is 1.66 dB, with linear variations from SLN = 1 to SLN = 31. Signal levels higher than −25 dBm are reported as SLN = 31. Signal levels lower than −75 dBm are reported as SLN = 0. These values correspond to isotropic reception antennas and may be modified if a directional antenna is used. The minimum defer threshold is −75 dBm, corresponding to SLN = 1. This is considered sufficient for burst transmissions from other HIPERLAN equipment. The channel would be idle only when SLN = 0. However, to cope with interference from other non-HIPERLAN equipment, whose transmissions are considered to be constant (not burst), an

120


adaptive threshold can be used. This adaptive threshold is calculated to be just above the constant interference level, as a result of the other equipment. It is not mandatory for HIPERLAN equipment to be able to calculate and maintain the adaptive threshold. The adaptive threshold must always be lower than or equal to the maximum adaptive defer threshold (MADT). The highest possible value of MADT is −40 dBm, corresponding to SLN = 22. A 5-ms measuring window is used to determine the threshold. A measurement can be taken at any time, including bursts of HIPERLAN transmission. Remember that the maximum length of a HIPERLAN burst is 1 ms. The SLN is evaluated during this window. The lower value of SLN in the 5-ms window is calculated, and the MADT is determined as two levels higher than this minimum SLN (never higher than SLN = 22). Figure 4.5 shows the calculation of the threshold. This MADT is used during the following 100 ms. Then the MADT is decreased by one SLN for the next 100 ms. For every subsequent period of 100 ms the MADT is decreased by one unit, until it reaches the minimum value of MADT = 1, or a new measuring window occurs. Figure 4.6 shows the aging process.

Signal level number

New threshold

n+2 n

Minimum SLN in measuring window

5 ms measurement window Figure 4.5

Calculation of the MADT.

Time


MADT Measurement window (5 ms)

121

Measurement window (5 ms) m

n

n-1

n-2

n-3

m-1

Data burst transmission (1 ms)

m-2

k k-1 k-2

100 ms Figure 4.6

100 ms

Time

Aging of the adaptive defer threshold.

After the successful transmission of an LBR-HBR data burst, and if a defer threshold different from SLN = 1 is being used, the equipment will measure the signal strength during the 256-bit periods (≈10 µs) immediately following the end of the burst. If the measured SLN is lower than the MADT − 2, then the new MADT is reduced to two levels higher than that of the SLN.

4.3

HIPERLAN channel access control (CAC)

4.3.1

Generalities

The CAC sublayer governs the establishment of the communications on a shared radio channel with the following objectives. ω

Allowing the user to establish priorities for the traffic;

ω

Preventing low-priority traffic from disturbing that of the highest priority;

ω

Freeing the user from the need to know the peculiarities of the communication channel;

122


Reducing the probability of collisions to an acceptable minimum, even in the presence of hidden nodes;

ω

Allowing a reasonable use of the channel.

HIPERLAN CAC is responsible for the following: ω

Selection of the radio channel;

ω

Implementation of the multiple access protocol, known as the EY-NPMA;

ω

Formatting the HCPDUs in transmission, including, if necessary, the data in the HMPDUs and transferring them to the PHY once the channel access is guaranteed;

ω

Receiving from the PHY the HCPDUs and decoding the corresponding data to be transferred to the MAC layer in the form of HMPDUs.

4.3.2

HIPERLAN CAC protocol data units (HCPDUs)

HCPDUs are exchanged between HC-entities. There are two types of HCPDUs: ω

The LBR-HCPDU is made up of a sequence of bits that is transferred at the LBR.

ω

The LBR-HBR-HCPDU is made up of an LBR-part and an HBR-part. The LBR-part consists of a sequence of bits that is transferred at the LBR. The HBR-part consists of a sequence of octets (groups of eight bits) that is transferred at the HBR.

There is only one LBR-HCPDU defined at the moment: the AK-HCPDU, which is used to acknowledge the reception of a unicast communication by its destined user. Figure 4.7 shows the structure of the AK-HCPDU. The first 10 bits of the AK-HCPDU are common to all HCPDUs, of any type. Bit 10 is the HBR-part indicator (HI) field. A 1 indicates that the HCPDU contains an HBR part. For an LBR-HCPDU, this bit is set at 0. The next eight bits form the ACK identifier field (AID). AIDCS is a four-bit field that carries a checksum on the AID. There are two LBR-HBR-HCPDUs defined in HIPERLAN. The channel permission HCPDU (CP-HCPDU) is used to interchange information on the permission to use channels 3 and 4, which are not allowed in all countries. This is needed, because HIPERLAN terminals can be moved from one network to another.


1010101001 10 bits (0-9)

Figure 4.7

123

HI 1 bit (10)

AID 8 bits (11-18)

AIDCS 4 bits (19-22)

Structure of the AK-HCPDU.

The data HCPDU (DT-HCPDU) is used to exchange all types of data between the users. The interchanged data come from the MAC and, eventually, from the applications. The structure of the CP-HCPDU is shown in Figure 4.8. The first 10-bit field is common to all HCPDUs. The second field (bit 10) is set to 1 to indicate that the HCPDU contains an HBR-part. The HBR-part is made up of blocks of 52 octets (416 bits). The number of blocks, n (between 1 and 47), is coded with six bits in two fields: the block length indicator (BLI) in the HBR part and the BLI replica (BLIR) in the LBR part. The field BLIR checksum (BLIRCS) is a four-bit checksum of the BLIR. The hashed destination is coded into the hashed destination address (HDA) field, and a checksum on the HDA is included in the HDACS. The HBR-part is made up of octets. The CP-HCPDU consists of only one block of 52 octets, 29 of them being padding octets (octets 20 to 48). The first octet carries the HCPDU-type indicator (two bits) plus the BLI (six bits). The second one is the padding length indicator (PLI), which is

1010101001 10 bits (0-9)

HI 1 bit (10)

HDA HDACS BLIR BLIRCS 9 bits 4 bits 6 bits 4 bits (11-19) (20-23) (24-29) (30-33)

1 1 bit (34)

A) LBR-part

TI&BLI 1 octet

PLI HID DA SA Ch. Inf. 1 octet 4 octets 6 octets 6 octets 1 octet

PAD CS 29 4 octets octets

B) HBR-part Figure 4.8

Structure of the CP-HCPCU: (a) LBR-part and (b) HBR-part.

124


set at 29 in this HCPDU. The HIPERLAN identifier (HID) field is made up of four octets (36) and the source and destination HCSAP addresses are included in two six-octet fields (DA and SA). Octet 19 carries the relevant information on the authorizations for channels 3 and 4. A 1 in bit 8 of this octet means that channel 3 is allowed. A 1 in bit 7 means that channel 4 is allowed. The rest of the bits in this octet are reserved for future use. Finally, the last four octets form the checksum field (CS), which carries a 32-bit checksum on the HBR-part of the LBR-HBR-HCPDU. The HBR-part of the DT-HCPDU structure is shown in Figure 4.9. The structure of the LBR-part is common to all LBR-HBR-HCPDUs, and thus it is equal to the one explained for the CP-HCPDU. Octets 1 to 18 of the HBR part are the same as in the CP-HCPDU. The user data field (UD) can carry any number of octets from 1 to 2,422 and contain the HIPERLAN CAC service data unit (HCSDU), the data for which the CAC service has been required by the MAC. The padding field contains 051 padding octets, so that the HBR part contains an integer number of blocks of 52 octets. Finally, the CS field is a 32-bit checksum calculated on the whole HBR-part. 4.3.3

Channel access

Channel access is attempted by the CAC entity according to the condition of the channel. The three possible channel conditions are described as follows: ω

Free channel: When the channel is detected as idle for a sufficiently large interval of time, access can be readily attempted.

ω

Synchronized channel: When the channel is busy at times, access can be made only at certain times according to the EY-CPMA protocol.

PLI HID DA TI&BLI SA 1 octet 1 octet 4 octets 6 octets 6 octets Figure 4.9

UD 1-2422 octets

Structure of the DT-HCPDU (HBR-part).

PAD CS 0-51 4 octets octets

The HIPERLAN Standard ω

125

Hidden elimination: As the HIPERLAN range is larger than the radio range of individual nodes, a situation may arise where two transmitters that are out of radio range of each other transmit data addressed to a third node that is in the radio range of both simultaneously, with the consequent collision. This situation is shown in Figure 4.10. The two transmitters are said to be nodes hidden from each other. The EY-CPMA protocol includes the means to detect the presence of hidden nodes transmitting. When a hidden node is detected, channel access is suspended for a certain time interval to prevent collisions.

The CAC protocol relies on three factors to avoid collisions: ω

Priority: Higher priority traffic has more immediate channel access. The protocol eliminates lower priority traffic from contention.

ω

Random yielding: The probability of equal priority traffic collisions is reduced by a yield phase that gives them equal access rights on average but that assigns random priorities to a particular transmission cycle.

Node C

Node A Radio Range of Node A

Figure 4.10

Hidden nodes.

Node B Radio Range of Node B

126


Detection of hidden transmitting nodes and transmission deferral for a short while to avoid collisions.

However, some collisions are likely. To save the integrity of the data, an acknowledgment is sent when a unicast transmission has been correctly received. This is not sent with a multicast transmission. 4.3.4

Channel access in the free channel condition

Transmission can be attempted after the channel is sensed as idle for a long enough interval. This is called the free channel interval, and its length is not equal for all nodes to avoid the possibility of collisions when several nodes detect that the channel is free at the same time. The duration of the channel-free interval is: i MF + n × i FS

(4.1)

where iMF is the minimum duration of the channel free interval. This is equal to 2,000 Th, Th being the HBR bit period (≈ 42.5 ns). Thus, iMF ≈85 µs. iFS is the duration of a slot of dynamic extension in the channel-free interval. This is equal to 212 Th, approximately 9 µs. n is a random number, with uniform probability between 0 and mFS. In the implementations mFS = 12. The channel-free interval length is between 85 and 193 µs. Different nodes that have been detecting that the channel is idle during the first 85 µs would attempt their transmission at different times, depending on the value of n that each one is using. The node with the smaller n would transmit first. The rest would detect that the channel is busy and wait for a new whole channel-free interval or attempt the transmission when the channel condition is synchronized. There is still a small likelihood that two nodes have the same random number n and transmit at the same time, with the consequent collision and loss of traffic. On average, 1/12 of the expecting nodes would have the same random number n. Figure 4.11 illustrates the situation in which two users attempt to use the channel once the channel-free condition is detected. As user 1 has a longer channel-free interval, it desists after detection of the transmission of user 2.


127

User 0 Channel free interval User 1 Channel free interval User 2 Channel free interval Transmission User K

Channel Idle

User 1 Transmits

Time Figure 4.11

4.3.5

Transmission in the channel-free condition.

Channel access in the synchronized channel condition

When in the synchronized channel condition, a channel access cycle is formed to reduce the probability of collisions. This cycle is made up of three phases: ω

The prioritization phase: This phase is intended to eliminate all the traffic except that with higher priority from contending the access to the channel within this cycle.

ω

The contention phase: In this phase, equal priority traffic contends for transmission within the cycle.

ω

The transmission phase: Nodes that have survived the elimination processes of the previous cycles attempt the transmission. When unicast transmission is made, this cycle ends with the reception (or absence) of an ACK.

At least one node survives the transmission phase. On some occasions, more than one node can survive the transmission phase, and thus collisions may occur. The protocol is specified to limit the likelihood of these collisions. Starting a channel access cycle The starting time of a channel access cycle is the end of a transmission, including the ACK. That means the following:

128


When a multicast transmission is made, the end of this transmission is the start of the following cycle, as no ACK is due.

ω

When a unicast transmission is made, the end of the corresponding ACK is the start of the following cycle, even in the case where the ACK is not received.

Prioritization phase After the start of the access cycle, the node must detect the channel for a number of prioritization slot intervals equal to their priority. Thus, nodes that have to send traffic with priority 0 may immediately access the channel with a channel access burst. Nodes with priority 1 must sense the channel for one slot, nodes with priority 2 must wait for 2 slots, etc. The nodes send a channel access burst only if they detect that the channel is idle for all the previous slots. Thus, a node with traffic with priority 1 would only access the channel if it did not detect any transmission with priority 0. Five priority levels are defined, from 0 to 4. The duration of the prioritization slot is 168 Th ≈ 7 µs. Figure 4.12 illustrates this phase. Elimination phase After the prioritization phase, several (or at least one) nodes will survive the competition. In fact, all nodes with the highest

Channel Access Burst (Priority 0) Channel Access Burst (Priority 1) Channel Access Burst (Priority 2)

Start of access cycle

Prioritization slot interval

Time

Figure 4.12 Prioritization phase. (Channel access bursts are sent only if in the previous slots the channel is idleor if there are not users with higher priority demanding transmission.)


129

priority among the first contenders will survive. To reduce the number of nodes a random procedure is performed in this elimination phase. The channel access burst minimum duration is 168 Th, equal to the duration of the prioritization slot. However, this is incremented by a random number of elimination slot intervals of 212 Th (≈ 9 µs). This number can be between 0 and 12. After transmitting in a slot interval, the probability of transmitting in the following one is 1/2. Thus, half the contending nodes (with equal priority traffic) are eliminated on average in each subsequent elimination slot. After a user stops transmitting the channel access burst, it must check the channel: If it is busy the user withdraws. If it is idle it goes to the yielding phase. In the elimination phase only the transmitters with longer channel access bursts survive. Again, there can be one or more contenders that survive this phase. The survivors must detect whether the channel will be idle for the following 256 Th after terminating their channel access burst, before proceeding to the next phase. Figure 4.13 illustrates the elimination phase. Yielding phase Once the elimination phase is finished, the surviving nodes are asked to yield for a random time before the start of transmission. If during the yield interval a user detects that the channel is busy, it withdraws. Thus, the user with the shorter yielding time is the first to transmit and catch the channel.

User 0 Access Burst

User 1 Access Burst

User 2 Access Burst

User 3 Access Burst

Start of Channel Access Burst (users with same priority)

Figure 4.13

Elimination phase.

Users 1 & 2 survive to the yielding phase

Time

130


Yield slot intervals are 168 Th long. A user must yield for a number, n, of yield slot intervals. n is a random variable with uniform distribution between 0 and 9. Thus, users with n = 0 would be the first to transmit, just at the end of the elimination phase. If there are no users with this n, then users with n = 1 would transmit, and so on. Figure 4.14 illustrates the yielding phase. After the successful reception of a unicast transmission, the destination node must release an AK-HCPDU. The start of it must be 512 Th (21 µs) after the end of the reception. The channel access cycle finishes after the ACK is received or its time interval has passed without sensing the ACK. This is to prevent a situation in which the ACK is sent by a hidden node. Figure 4.15 shows the whole channel access cycle in the synchronized channel condition. Acknowledgment

4.3.6

Hidden node detection and operation

When, after losing its right to transmission in the synchronized channel operation, the node does not detect data transmission, it assumes the presence of a hidden node and enters the hidden elimination condition for the following 500 ms. When in this condition, there are some channel access suspension intervals, made up of a number of slots of 1 ms (one to five slots, with equal probability) during which the channel access attempts are suspended. These suspension intervals start at the following times:

Yield Interval User 1

Transmission User 1

Yield Interval User 2

User 1 transmits User 2 withdraws

Start of yielding phase Figure 4.14

Yielding phase.

Time


131

Multicast transmission Unicast transmission Unicast transmission

ACK Missing ACK

User 1 Access Burst User 2 Access Burst

Yield Interval User 1 Yield Interval User 2

User 1 transmits

Yielding phase

Transmission phase

User 2 withdraws

User 3 Access Burst Previous cycle Prioritization transmission phase phase

Elimination phase

Start of Access Cycle

Figure 4.15

Synchronized channel access cycle.

ω

Every time the hidden node condition is detected.

ω

When, being in the hidden elimination condition, an LBR-HBRHCPDU transmission is detected, but the corresponding ACK is missed. Next, the medium access control of HIPERLAN is presented.

4.4

HIPERLAN medium access control (MAC)

4.4.1

HIPERLAN MAC functions

HIPERLAN MAC performs several functions that are part of the DLL in the OSI model. Among these functions, it does the following: ω

Provides the transmission of HMPDUs using the CAC services, to one or several destinations within the same HIPERLAN, with or without data encryption. This can be carried out with regard to the time available for traffic.

ω

Implements the feature of HIPERLAN regarding the network range larger than the radio range of individual nodes, with the aid of forwarding nodes acting as relaying stations that deliver to other nodes the data that are addressed to them.

ω

Maintains dynamic information regarding other stations within the same network. This information includes HIPERLAN identification and name, nodes and distance to reach them with indications on the best route, etc. All this information is updated dynamically.

132


Allows the implementation of power-saving functions that are intended to reduce the stations power consumption, particularly those that are battery-operated.

HIPERLAN nodes can be forwarder or nonforwarder. Forwarder nodes may retransmit traffic received from their neighboring nodes (those that are within their radio range), acting as relaying stations. A local node must select from its neighboring forwarder nodes those that will act as multipoint relays for it. The selected forwarder nodes must retransmit the multicast traffic originating at the local node and the traffic addressed to destinations that the local node has selected to route through this particular forwarder node. 4.4.2

HIPERLAN differentiation and addressing

Because of the open nature of the radio medium, HIPERLANs must have some identification to differentiate them from others that may share the radio channel. There are two parameters for this differentiation: ω

HIPERLAN name: This is made up of 32 characters of 16 bits;

ω

HIPERLAN identifier: This is a numerical field with 32 bits.

The impossibility of maintaining global HIPERLAN name and identifier registrations, so that each user has a unique identification, raises the possibility of two overlapping HIPERLANs having the same name and identifier and thus mingling their traffic. However, the probability is so small that it is very unlikely. To ensure further traffic protection, data encryption may be an option. The HIPERLAN identifier, 0, is reserved for Any-HIPERLAN, which makes reference to any network within the radio range and is used in some functions that will be described later. HIPERLAN identifiers with the highest order bit set to 1 are used for HIPERLANs that make use of data encryption. If the highest order bit is set to 0 the HIPERLAN does not use data encryption. HIPERLAN uses the 48-bit LAN MAC address established in the ISO MAC service definition. There are individual addresses and group addresses. The former may be used as source as well as destination addresses. Group addresses are only used as destination addresses. For practical reasons, the same 48-bit address is used to identify the HIPERLAN MAC service user (HMS-user) attached to the MSAP, the


133

MSAP itself, the HIPERLAN MAC entity, the attached HIPERLAN CAC entity, and the HCSAP. This is valid for individual or group addresses. 4.4.3

Data encryption

User data may be encrypted by mixing it (with an XOR operation) with a pseudorandom sequence. This sequence is defined through a user key and an initialization vector. The method of generating the pseudorandom sequence is not public and is available from ETSI for authorized users. Both the key and the initialization vector have 30 bits. Three keys can be used simultaneously within a net, being identified by a key identifier. This field has the value 0 (no key) or 13 to select one of them. Figure 4.16 shows the HIPERLAN encryption-decryption scheme. Encryption is only performed on the user data. The rest of the fields in the HMPDUs are transmitted as clear information to be available to every node. 4.4.4

Power-saving function

To reduce power consumption, algorithms are provided within the HIPERLAN standard. HIPERLAN nodes can operate as the following: ω

P-savers: Nodes that try to reduce power consumption by having their receivers active only during some intervals;

Key identifier

Key identifier

Key identifier

Key-set

Key-set Initialization vector

Random sequence generator

XOR

Initialization vector

Encrypted data

Initialization vector

XOR

Transmitted data Figure 4.16

HIPERLAN encryption-decryption scheme.

Random sequence generator

Received data

134


P-supporters: Nodes that adapt their transmissions to the active intervals of p-savers.

HIPERLAN nodes can act as p-savers or p-supporters or as both or neither of them. The user must define the ways of operation of this function. Regarding individual traffic (with only one destination), p-savers must declare their attention interval. This interval is defined within three parameters, according to Figure 4.17. The parameters are described as follows: ω

Pattern period: The period of repetition of the active intervals;

ω

Practice interval: Duration of the active intervals;

ω

Pattern offset: A time reference for the start of the current pattern period relative to the declaration of the pattern.

P-savers regularly declare their individual attention pattern. P-supporters adapt their transmissions destined to a p-saver to their practice interval. However, regarding group-addressed traffic, this is not enough. For this, p-supporters define and regularly declare their groups attention pattern. This has the same structure as previously shown in Figure 4.17 and the same parameters. All groups destined traffic is delivered within the practice interval. Pattern Offset Pattern Period

Practice Interval

Practice Interval Time Pattern Declaration

Figure 4.17

Individual or group attendance pattern.


135

Although the standard does not elaborate more on this aspect, it is clear that in a practical network some kind of organization is needed so that traffic is not lost and so that a large number of retransmissions are not needed. Some examples follow: ω

In a network with a mains-operated node and a number of portable battery-operated nodes, the latter could act as p-savers and the former as p-supporters. Each portable node would declare its individual attention pattern, and the p-supporter should declare its group attention pattern. The portable nodes should still be active for their own individual practice interval but also for the group practice interval.

ω

A better solution would be to synchronize them, so that all portable nodes would declare individual attention patterns equal to the group attendance pattern. This would save battery power further. All transmissions from the p-supporter, be they individual or group addressed, would take place during the common interval.

ω

However, the traffic directed from one portable node to another would be sent at any time and thus lost except in the practice interval of the destination user. The lack of an acknowledgment would help, but a lot of retransmissions would be needed. A better solution would be for the portable nodes to act as p-supporters to their neighbors, thus adopting their transmissions to their practice interval.

Were all the nodes p-supporters and p-savers, with all their practice intervals synchronized, the consequence would be a net that is active only a part of the time, during the practice intervals of all the nodes. This could suffice for low levels of traffic but would affect the network capacity during traffic peaks. Thus, a means to adapt the practice interval length, or even disconnect the power-saving functions during traffic peaks, would be needed to avoid compromising the network traffic capacity. As with many other features, it is up to the manufacturers and users to make decisions on the ways in which HIPERLAN facilities are operated. 4.4.5

MAC information databases

The functionality of HIPERLAN MAC requires the maintenance of several databases; information in the databases is dynamically updated on

136


the basis of data received from the neighboring nodes. All the information has a holding time and is removed after this time has passed. The databases are described as follows. 1. P-saver information base: Each p-supporter maintains a database with the individual attention pattern of its neighboring p-savers. These data are refreshed after a p-saver declaration of its pattern, and has a default holding time of 30s. After this time it is removed. 2. P-supporter information base: Each p-saver has a base with the group attention pattern of its neighboring p-supporters. These data are refreshed when a group attention declaration is received. It has a default holding time of 30s. 3. Duplicate detection information base: It is possible for a HIPERLAN node to receive information that it has previously transmitted or received. This is particularly true with multicast HMPDUs that are received by the same node that has transmitted them and that are retransmitted by the neighboring forwarding nodes. To detect this duplicate information, a base with the HMPDU sequence number and source address is maintained, and both parameters of a received HMPDU are checked to discard them in case they are in the database. The information is held for the residual lifetime of the data. This residual lifetime is the time beyond which the information is useless and is no longer transmitted. 4. Route information base: This information is needed to route the traffic to nodes that are outside the radio range of the local node. The base includes data on the neighboring forwarder node through which the destination can be reached and the estimated number of hops to reach it. 5. Neighbor information base: This specifies for each of the neighboring nodes their status with respect to the local node. This status may be asymmetric (if the link is asymmetric), symmetric, or multirelay (if the node is selected to be the multipoint relay of the local node). 6. Hello information base: This base records the information provided by the neighboring nodes. It has three fields: the address,


137

status, and the node to reach it. The status can be neighbor forwarder, neighbor nonforwarder, and two-hopif it has a mutual neighbor forwarder with the local node. 7. Source multipoint relay information base: A node that has been selected as the multipoint relay of a neighboring node must keep this information within this base. It stores the address of the source multipoint relay and the sequence number assigned. (Each multipoint relay set has a sequence number that is incremented each time the set is updated.) 8. Topology information base: This base stores information on the source multipoint relays of other forwarders. It is made with entries with three fields: destination address, multipoint relay address, and sequence number. The node specified in the destination address can be reached with one hop from the multipoint relay address. 9. Alias information base: Destinations outside the HIPERLAN are assigned an alias address within the network to apply routing procedures. Not all the nodes must maintain the nine databases. The first two bases are needed to implement the power-saving functions that are optional to the nodes. The third one is intended to detect and remove duplicated information, and the other six are in relation to the routing procedures. Forwarder nodes must keep the six databases, but nonforwarder nodes keep neither the source multipoint relay information base nor the topology information base. 4.4.6

Priorities and traffic lifetime

When the user requests the transmission of a data unit, the MAC is ready to provide a connectionless transmission service. This means that the transmission is made at once, without an establishment process or a communications release. Each data unit is managed independently of any other and can be transmitted in any order. An indication of order is set as a sequence number, so that the destination node can rearrange the received data units to their original order. Because of the lack of information in the network on data flow and delays and the dynamic characteristics of the network topology, power-

138


saving functions, and routing information, no guarantee can be made to the user as to the order of transmission, ways of reaching the destination, and the time needed to reach it. This is no problem with some types of communications that can be considered asynchronous. For example, the transfer of files between computers does not usually have a time interval within which all data has to be received, and the order in which information is received is not important as long as the original order can be restored. HIPERLAN allows also the transmission of time-bounded communications, data that have to be delivered within some time interval and that is useless outside it. For example, in voice communications it would not make sense to stop the decoding process if one packet of data were late. Real-time applications are also sensitive to delays and thus need an assessment of the time of delivery. When the user requests the transmission of a data unit, it must specify the local priority (0 = urgent, 1 = default) and the lifetime of the unit. In the transmission process a residual lifetime is kept so that the channel priority is set according to the user priority and the residual lifetime. Both the initial and the residual lifetime can range from 0 to 16,000 milliseconds. The default value of the initial lifetime is 500 ms. Channel priorities, ranging from 0 (highest priority) to 4 (lowest priority) are set according to Table 4.2. Normalized residual lifetime (NRL) is the residual lifetime divided by the estimated number of hops to reach the destination. As described in Section 4.3, the highest priority traffic is sent before considering the transmission of traffic with lower level priority. In a well-loaded network, in peaks, only traffic with priority 0 or 1 would be sent, but all the traffic would reach these priorities when approaching the Table 4.2 Channel Priority Establishment NRL

User User priority 0 priority 1

NRL < 10 ms

0

1

10 ms ø NRL < 20 ms

1

2


2

3


3

4

80 ms ø NRL

4

4


139

limit of their lifetime. Traffic that cannot be sent before the extinction of its lifetime is discarded, and an indication of failure to the user is made. 4.4.7

Types of HMPDUs

There are seven types of HMPDUs: ω

Type 1: DT-HMPDU;

ω

Type 2: Look-up request HMPDU (LR-HMPDU);

ω

Type 3: Look-up confirm HMPDU (LC-HMPDU);

ω

Type 4: Individual-attention pattern HMPDU (IP-HMPDU);

ω

Type 5: Group-attention pattern HMPDU (GP-HMPDU);

ω

Type 6: Topology control HMPDU (TC-HMPDU);

ω

Type 7: Hello HMPDU (HO-HMPDU).

DT-HMPDUs are used to transmit the user data. All the other types are needed for functions of the HIPERLAN. LR-HMPDU and LC-HMPDU are used in connection with the look-up function, by which the node may discover the HIPERLANs within its range by requesting information from their neighboring nodes. IP-HMPDU and GP-HMPDU are used to declare the attention patterns of p-savers and p-supporters, in the power-saving functions. Finally, TC-HMPDU and HO-HMPDU are used to interchange information needed for routing functions. HMPDUs have a minimum of three and a maximum of 2,422 octets. The first two octets are the field length indicator (LI), which codes with a binary number the total number of octets in the HMPDU. The third one is always the type indicator, which encodes with a binary number the type (1 to 7) of HMPDU. 4.4.8

Look-up function: LR-HMPDU and LC-HMPDU

When a node wants to know which HIPERLANs are within its range, it sends a look-up request. This is sent with an LR-HMPDU, whose structure is shown in Figure 4.18. This is the shortest HMPDU, with only the first three octets being mandatory. The third one is set at 2, coding an LR-HMPDU. When a node receives an LR-HMPDU, it sends a look-up confirmation, using an LC-HMPDU, whose structure is shown in Figure 4.19.

140


LI=3 2 octets Figure 4.18

Structure of the LR-HMPDU.

LI=71 2 octets

Figure 4.19

TI=2 1 octet

TI=3 1 octet

HID 4 octets

HN 64 octets

Structure of the LC-HMPDU.

The HID and HN fields contain the HIPERLAN identifier (32 bits) and name (32 characters of 16 bits). Upon reception of one or several LC-HMPDUs, the node knows which HIPERLANs it can reach via radio. 4.4.9

IP-HMPDU and GP-HMPDU

P-savers and p-supporters must declare their patterns to implement the power-saving and attendance pattern functions. The declarations are made with the IP-HMPDU for the individual attendance pattern of p-savers and the GP-HMPDU for the group attendance pattern of p-supporters. Figure 4.20 shows their structure. The type indicator is 4 or 5 for the IP-HMPDU or the GP-HMPDU. The pattern offset, pattern period, and practice interval are declared as binary 16-bit numbers. They represent the intervals in milliseconds, with a maximum of 10,000. Pattern periods and practice intervals have a minimum of 500 ms. 4.4.10

DT-HMPDU

The DT-HMPDU has to transport the user data coming from the MSDU. Its structure is shown in Figure 4.21. The first two fields are common to all HMPDUs. The type indicator is set at 1. Residual lifetime (RL), in milliseconds, is coded into field RL, with two octets. The sequence number is coded with two octets in field PSN.


LI=9 2 octets

Figure 4.20

LI 2 octets

TI=1 1 octet

Figure 4.21

TI=4 or 5 1 octet

141

PO 2 octets

PP 2 octets

PI 2 octets

Structure of the IP-HMPDU and GP-HMPDU.

RL DA ADA PSN ASA UP&ML KID&IV SA 2 octets 2 octets 6 octets 6 octets 6 octets 6 octets 2 octets 4 octets

UD Variable length

SC 2 octets

Structure of the DT-HMPDU.

Because the order in which data units are received can be different from the order of transmission, the sequence number can be used to rearrange them. Four six-octet fields are included to carry the source address (SA) and destination address (DA) and alias source address (ASA) and alias destination address (ADA). Octets 32 and 33 carry the user priority (UP) field (one bit) and the MSDU lifetime (ML) (15 bits). Octets 34 to 37 carry the information on the data encryption. The first two bits are the key identifier field (KID), with values 0 (no key) to 3, and the 30 bits carry the initialization vector (IV). User data of variable length goes, encrypted or not, into the UD field. The sanity check (SC) field is used to take care of possible errors. The two octets are set to zero when encryption is not employed. Otherwise, the contents of SC are calculated using the data in the fields KID, IV, and UD. 4.4.11

TC-HMPDU and HO-HMPDU

Both HMPDUs are used to exchange information on the network structure, for routing purposes. The TC-HMPDU is used by forwarder nodes to declare its information on neighbors selecting it as multipoint relay. HOHMPDUs are sent by all nodes and contain information on the node neighbors. Figure 4.22 shows the TC-HMPDU structure. The first fields (LI, TI, RL, and PSN) have the previously defined meaning. There is the originator address (OA) field; the rest of the HMPDU is made up of pairs of multipoint relay set sequence number (MSN) fields and source multipoint relay address (SMA) fields.

142


RL LI TI=6 PSN OA MSN SMA {MSN,SMA} Pairs 2 octets 1 octet 2 octets 2 octets 6 octets 2 octets 6 octets 8 octets each Figure 4.22

Structure of the TC-HMPDU.

The MSN and SMA information is stored in the source multipoint relay information database of the originating node, which is a forwarder node. Forwarder nodes that receive the TC-HMPDU save the information in their topology information base and use it for routing purposes. All addresses are coded in a six-octet field. The sequence numbers are stored in a two-octet field. Figure 4.23 shows the greeting information HMPDU (HO-HMPDU) structure. The first two fields are set up as usual. The type indicator is set to 7. The third field (RTI) contains the relay type indicator (RTI) of the originator. The fourth is the MSN. Every time the node updates its list of neighbors selected as multipoint relays, it increases by one this sequence number. The rest of the HMPDU is made up of pairs of neighbor addresses (NAs) and neighbor status (NS). Addresses are codified with six octets and status with one octet. This is the information contained in the originators neighboring information base. 4.4.12

Routing functions and information maintenance

HIPERLANs are dynamic in nature, because some nodes are being added or removed (for example, if a portable computer is switched on and off), but also because of the mobility allowed to terminals and the irregular characteristics of the propagation medium. Some situations that can occur in HIPERLANs that are not possible in wired networks are described as follows:

LI TI=7 2 octets 1 octet

Figure 4.23

RTI NA MSN NS 1 octet 2 octets 6 octets 1 octet

Structure of the HO-HMPDU.

{NA,NS} Pairs 7 octets each


143

ω

The remotion of a forwarder node can fragment the HIPERLAN, as some nodes that could be reached through it may not be reachable by other means.

ω

The remotion of a forwarder can enforce a change in the topology, in the sense that some nodes are still within the radio range of the HIPERLAN, but the route to reach them has changed.

ω

A node can change its position, and thus the route to reach it may be different than before. As terminals can be mobile, this kind of change might be repeated frequently.

ω

Propagation conditions can change, and some terminals that were in their mutual radio range can no longer be reachable from another. This is particularly true in the indoor environment, where even the movement of people can affect the propagation channel.

Apart from the obvious situation where all nodes are in LOS of one another, the HIPERLAN nodes must keep updated references of the network situations. This is made using the previously described databases. These databases are continuously updated on the basis of information exchanged between the terminals in the form of HO-HMPDUs and TCHMPDUs. The information gathered from these HMPDUs has a relatively short holding time (20s and 40s). It is recommended that nodes send an HO-HMPDU and a TC-HMPDU (if they are forwarders) at intervals approximately equal to half the corresponding information holding time. Two nodes are said to be neighbors if they can communicate with each other directly via radio. Nodes that are not neighbors can be reached through forwarder nodes. If there is enough routing information in the local node, the transmission can be made more efficiently, using the shortest path to reach its destination. Otherwise, the information sent to a remote node can be transmitted to all forwarder neighbors, and they have to take care of delivering it to its destination. Local information in the databases is based on the information received from their neighbors. The full process is entirely dynamic: A new node entering the network would have empty databases. It would listen to the received HO-HMPDUs and TC-HMPDUs to insert data into its databases. The order in which this is made will be explained next.

144


The new node would transmit its own HO-HMPDUs and TCHMPDUs. The first ones would carry little or no information, but the local identifications would be useful to their neighboring nodes in identifying this new node. They would update their database to include it. In the meantime, the local databases would begin to contain information that would be communicated to the neighbors in the subsequent transmission of these HMPDUs. Shortly afterward, the new node would have complete information on the network, and all the other nodes would have updated their information to include it. After the removal of a node, all the information on it that is in its neighbors database would expire and thus be removed. As the content of these databases is transferred to the rest of the nodes, in little time all the information regarding the removed node would disappear from the network. The net would reconfigure itself to reach all of the nodes in case the removed node was a forwarder. As can be seen, this is a rather automatic process and completely decentralized. There is no node acting as server with all the information. The information is distributed between all the nodes. Let us review the information databases function, their role in the routing procedures, and their update processes: ω

All nodes keep four databases used for routing traffic: the alias information base (AIB), the neighbor information base (NIB), the hello information base (HIB), and the route information base (RIB). Forwarder nodes also keep two other databases: the topology information base (TIB) and the source multipoint relay information base (SMRIB).

ω

The AIB is updated every time a DT-HMPDU is received with a valid ASA. Its address and alias address are stored. This information is used to apply routing procedures (based on the alias address) to addresses outside the network.

ω

When an HMPDU has to be transmitted to an individual address, the route to reach the destination is obtained from the RIB. This includes entries with three fields: DA, forwarder through which the destination can be reached, and estimated number of hops. Thus, this is the most important database. The routing maintenance functions try to keep its information updated.


145

The RIB is updated each time one of the following databases is changed: ω

The NIB, which contains information on the neighbors in entries with two fields, a neighbor address and its status (symmetric, asymmetric, and multipoint relays). The neighbor status is asymmetric if the local node is not known by the neighbor, because it is not included in the HO-HMPDUs received from it. A neighbor is stated as multipoint relay if it is a forwarder and has been selected by the local node as the relay to reach some other nodes.

ω

The TIB, which exists only in forwarder nodes. It includes the information received in TC-HMPDUs sent by other forwarder nodes. This information comes from their SMRIBs.

Upon carrying out modifications on the NIB or the TIB (or both), the RIB is emptied, and new entries are registered. First, all the entries in the NIB whose status is not asymmetric are transcribed to the RIB as nodes reachable in one hop. Nodes that can be reached in two or more hops are obtained from the TIB. For each destination, the shortest route is selected and annotated in the RIB. The reception of a TC-HMPDU initiates the process of updating the TIB. The information received is annotated, and, if there are changes, the RIB is also updated. After the reception of an HO-HMPDU, three databases may be updated: ω

The originator of the HO-HMPDU is included in the NIB and its status updated.

ω

Entries in the HIB are generated to include the originator as neighbor, and the originators neighbors that are not local neighbors. These are recorded as nodes that can be reached in two hops.

ω

In case the local node is a forwarder and has been selected as multipoint relay by the originator of the HO-HMPDU, the SMRIB is updated to include this information.

If there are changes in the NIB, then the RIB is also updated. The NIB is also updated when a multipoint relay selection procedure is performed. This selection is made on the basis of the information in the NIB and the HIB. For each entry in the HIB that can be reached in two hops a

146


multipoint relay is selected from the forwarders in the NIB. If the selection is different from the previous one, the sequence number is increased, and the NIB is accordingly modified. Then the RIB is also updated. To summarize, nodes keep a number of databases with information useful for routing traffic. The information is based on data received from other nodes, in the form of HO-HMPDUs and TC-HMPDUs, and on the multipoint relay selection that nodes must perform at certain times. All nodes must regularly transmit HO-HMPDUs and TC-HMPDUs (if they are forwarder nodes).

4.5

Conclusions on HIPERLAN type 1

The European standard HIPERLAN allows the implementation of WLANs based on standard equipment with highly advanced functionality. The main features of HIPERLAN that make it attractive when compared to other WLANs can be summarized as follows: ω

An HBR of 23 Mbps, comparable to or even higher than that of a WLAN;

ω

Virtually unlimited coverage, as it does not depend on the radio range of individual nodes, thanks to the forwarding and routing functions and an easy interconnectability with other networks;

ω

Dynamic configuration of the network, with a fast update upon making changes in the nodes;

ω

High traffic capacity, the means to support time-bounded as well as asynchronous traffic;

ω

Data encryption and power-saving functions.

The main disadvantages are related to the use of the 5-GHz band, instead of the more common 2.5-GHz band. As the technology is not as mature in that band (the 2.5-GHz band has been used for unlicensed operation for more than 10 years), HIPERLAN products may take some more time to reach the market and may initially be more expensive. However, there is no reason why the mass production costs may not reach the same low levels as those of 2.5-GHz products in a short while.


4.6

147

Future BRAN standards

The previous sections extensively describe the basic HIPERLAN type 1. From 1990 to 1992, CEPT designed three bands for allocating new wireless network technologies: the 2.4-GHz ISM band for wideband data systems using spread spectrum techniques and the 5.2- and 17.1-GHz bands for HIPERLAN. CEPT has reserved 100 MHz of the spectrum in the 5.2-GHz band, with a further 50 MHz available at the discretion of national administrations,1 and 200 MHz in the 17-GHz band [1]. EMC criteria for these bands are described in [4]. ETSI is developing three new BRAN standards to be allocated in these new bands. HIPERLAN type 2 (wireless IP, ATM, and UMTS short-range access) [5] is designed to provide local wireless access to ATM, IP, and UMTS infrastructure networks for both moving and stationary terminals that interact with the access points that are connected to the infrastructure networks. It will be able to provide the same QoS that users would expect from a wired IP or ATM network. The typical operating environment is indoor, with mobility restricted to the local service area. The data rate is on the order of 25 Mbps. The technical specification for the PHY was approved in October 1999. It is intended for the 5-GHz unlicensed band. HIPERACCESS (wireless IP and ATM remote access), previously known as HIPERLAN 3, will provide, high-speed, outdoor (25-Mbps) wireless access to infrastructure networks. Unless type 2, HIPERACCESS is not designed to support mobility at high data rates, allowing the use of directional antennas with significant gains. Thus, the range can be increased to 5 km. This will allow the rapid deployment of broadband access WANs. It will operate in the licensed (360-GHz) as well as the unlicensed (5-GHz) bands. The HIPERACCESS specification is expected to be completed within the first months of 2000. HIPERLINK (wireless broadband interconnect), previously known as HIPERLAN 4, is intended to provide point-to-point (up to 150m) veryhigh-speed wireless links. It will operate in the 17-GHz unlicensed band, with bit rates up to 155 Mbps. One of the envisaged applications is the interconnection of HIPERACCESS networks and/or HIPERLAN APs in a fully wireless network. Figure 4.24 presents an overview of the different

1. Because of its current level of occupation, the 5.2-GHz band is expected not to be sufficient for projected HIPERLAN needs. ETSI and CEPT are discussing further spectrum assignations.

148

Wireless LAN Standards and Applications HIPERLAN

HIPERLAN

HIPERACCESS

HIPERLINK

Type 1 Wireless 8802 LAN

Type 2 Wireless IP, ATM and UMTS Short Range Access

Wireless IP and ATM Remote Access

Wireless Broadband Interconnect

MAC

DLC

DLC

DLC

PHY (5 GHZ) 19 Mb/s

PHY (5 GHZ) 25 Mb/s

PHY (various bands) 25 Mb/s

PHY (17 GHZ) 155 Mb/s

Figure 4.24

Overview of the different BRAN standards.

BRAN standards and their allocation and baud rates. HIPERLAN type 2, HIPERLINK, and HIPERACCESS can be combined into an open wireless architecture that meets the needs of a very large user population. ETSI has described two main environments for this kind of network. The domestic premises network environment (DPN) covers the home and its immediate vicinity. It typically includes a localized radio extension to a broadband network. It is characterized by individual cells, and supporting mobility beyond the coverage area is not required. The business premises network (BPN) environment covers a privately owned network, over an extended area (such as university campuses or hospitals). It may offer access switching and management functions within an arbitrarily large coverage area serviced by multicellular wireless communications facilities. We can define some application scenarios within these environments for HIPERLAN 2, HIPERACCESS, and HIPERLINK, such as replacing infrastructure networks or wireless access to infrastructure networks in the DPN or interconnection of manufacturing devices in BPN. In BPN, delay and data losses are critical because of the need for supporting alarm data and other time-bound services.

References [1]

CEPT Recommendation T/R 22-06, Harmonized Radio Frequency Bands for High Performance Radio Local Area Networks (HIPERLANs) in the 5-GHz and 17-GHz Frequency Range, Madrid, 1992, revised at Nicosia 1994.


149

[2]

ETS 300 652 V.1.2.1, Broadband Radio Access Networks (BRAN); HIgh PErformance Radio Local Area Network (HIPERLAN) Type 1; Functional specification, European Telecommunications Standards Institute (ETSI), July 1998.

[3]

LaMaire, R. O., et al., Wireless LANs and Mobile Networking: Standards and Future Directions, IEEE Comm., Vol. 34, No. 8, Aug. 1996, pp. 8694.

[4]

ETS 300 826, Electromagnetic Compatibility and Radio Spectrum Matters (ERM); Electromagnetic Compatibility (EMC) standard for 2.4-GHz Wideband Transmission Systems and High Performance Radio Local Area Network (HIPERLAN) equipment, European Telecommunications Standards Institute (ETSI), Nov. 1997.

[5]

TR 101 031 V.2.2.1, Broadband Radio Access Networks (BRAN); HIgh PErformance Radio Local Area Network (HIPERLAN) type 2; Requirements and Architectures for Wireless Broadband Access, European Telecommunications Standards Institute (ETSI), Jan. 1999.

Selected Bibliography ETS 300 652/A2, Radio Equipment and Systems (RES); HIgh PErformance Radio Local Area Network (HIPERLAN) Type 1; Functional specification. Amendment 2, European Telecommunications Standards Institute (ETSI), Jun. 1998. ETS 300 836-1, Broadband Radio Access Networks (BRAN); HIgh PErformance Radio Local Area Network (HIPERLAN) Type 1; Conformal Testing Specification. Part 1: Radio Type Approval and Radio Frequency (RF) Conformance Test Specifications, European Telecommunications Standards Institute (ETSI), May 1998. ETS 300 836-2, Broadband Radio Access Networks (BRAN); HIgh PErformance Radio Local Area Network (HIPERLAN) Type 1; Conformal Testing Specification. Part 1: Protocol Implementation Conformance Statement (PICS) pro forma Specification, European Telecommunications Standards Institute (ETSI), May 1998. ETS 300 836-3, Broadband Radio Access Networks (BRAN); HIgh PErformance Radio Local Area Network (HIPERLAN) Type 1; Conformal Testing Specification. Part 3: Test Suite Structure and Test Purposes (TSS&TP) Specification, European Telecommunications Standards Institute (ETSI), May 1998. ETS 300 836-4, Broadband Radio Access Networks (BRAN); HIgh PErformance Radio Local Area Network (HIPERLAN) Type 1; Conformal Testing Specification. Part 4: Abstract Test Suite (ATS) Specification, European Telecommunications Standards Institute (ETSI), May 1998.

150


Halls, G., HIPERLANThe 20-Mbps Radio LAN, IEE Elect. Div. Colloquium on Radio LANs and MANs, IEE Colloquium Digest, No. 071, Stevenage, United Kingdom, 1995, pp. 1/11/8. Minnet, P., HIPERLAN: A European Standard for 5-GHz and 17-GHz Wireless LAN, Microwave Engineering Europe, Oct. 1995, pp. 6166. Pahlavan, K., A. Zahedi, and P. Krishnamurthy, Wideband Local Access: Wireless LAN and Wireless ATM, IEEE Comm., Vol. 35, No. 11, Nov. 1997, pp. 3440. TR 101 054, Security Algorithm Group of Experts (SAGE); Rules for the Management of the HIPERLAN Standard Encryption Algorithm (HSEA), Ed. 1.1.1, European Telecommunications Standards Institute (ETSI), Jun. 1997. TR 101 177, Broadband Radio Access Networks (BRAN). Requirements and Architectures for Broadband Fixed Radio Access Networks (HIPERACCESS), European Telecommunications Standards Institute (ETSI), May 1998. TR 101 378, Broadband Radio Access Networks (BRAN); Common ETSIATM Forum Reference Model for Wireless ATM Access Systems (WACS), European Telecommunications Standards Institute (ETSI), Dec. 1998. Wilkinson, T., T. G. C. Phipps, and S. K. Barton, A Report on HIPERLAN Standardization, International Journal on Wireless Information Networks, Vol. 2, Apr. 1995, pp. 99120.

CHAPTER

5

Contents 5.1

Introduction

5.2 The application scenarios 5.3 Wireless LAN technologies and products 5.4

Conclusions

Application Scenarios A. Santamaría, V. M. Melián, and J. C. Miñano

5.1

Introduction

Due to the evolution of wireless networks in recent years, there is now a wide range of commercial products that provide satisfactory solutions for the two main problems posed by wired networks: portability and reconfiguration. Typical applications of wireless networks can be classified into two categories: ω

Indoor applications;

ω

Building-to-building interconnections.

Indoor applications deal with the interconnection of LAN equipment inside a building using wireless links. Usually, buildings have a wired backbone LAN. To install a wireless network, AP equipment is connected to the backbone. Wireless user equipment is connected to the AP equipment through wireless links that can be both RF or 151

152


IR. The portability afforded by wireless links allows users to avoid reconfiguration or reinstallation costs. Users have total freedom and flexibility to place their equipment anywhere inside the covered area. Temporary work groups can be easily connected through a WLAN. Wireless networks are also used to extend rapidly wired networkswith a lower cost than wired extensions. In addition, they make it possible to develop an agile and easy-to-grow LAN architecture. The security features of wireless networks, as well as their interoperability and performance, are comparable to those of wired networks. However, a new application scenario arises with computers used at home. Home computer users can utilize wireless links to avoid installing wire along corridors or connecting the different rooms in a house. Today, several pieces of commercial equipment easily allow the installation of WLAN indoor links. This chapter briefly discusses the main application areas in which WLANs can be used, organizing the applications into four categories: public building WLANs, business environment WLANs, domestic building WLANs, and industrial sector WLANs. The chapter provides a short description of user and system needs given for each application category. Finally, the chapter details the available commercial products for WLAN, analyzing the markets for both technologies (IR and RF) and presenting their main characteristics. Companies that offer WLAN products have joined common trade and technological associations; Section 5.3 discusses these organizations, which include IrDA, WLANA, the Wireless Ethernet Compatibility Alliance (WECA), the Bluetooth Special Interest Group (Bluetooth SIG), the WLAN Interoperability Forum (WLIF), and the HomeRF. Next the application scenarios of WLANs are presented.

5.2

The application scenarios

5.2.1

Public buildings

In todays global environment, the development of intelligent buildings based on wireless communication networks has become important. The mobility of people combined with the need for an integrated communication system have stimulated the development of advanced indoor wireless communication systems. The public buildings sector imposes specific requirements due to the nature of the functionality of public buildings. This section presents the

Application Scenarios

153

scenarios for wireless communication networks in public sector buildings. The communications environment of public buildings is different from other wireless communications environments because of the presence of uninvited people. This presence creates new needs related to security, building functionality, and/or maintenance of the communication service. Some possible scenarios for wireless communications in public buildings are described as follows. ω

Education scenarios: Bearing in mind the development of IT, the students, teachers, and administration personnel of a school campus need campus-wide connectivity. Wireless solutions are cheaper to install than traditional wired solutions and easy to reinstall. Wireless links also provide high-speed communications. Several educational centers have tried out wireless solutions with satisfactory results. Some examples follow: ω

ω

Battle Creek, Michigan, schools: The Battle Creek Michigan school district, known for academic excellence, faced the challenge of providing Internet access to students and teachers from all five middle schools. Internet connectivity allows users to access the learning material available on the Web. This material is considered essential for classwork and teacher preparation. The Internet access service was planned based on LAN architecture. Because the Michigan schools are located in an urban environment that includes historic brick buildings, wire installation was too expensive. Accordingly, the school system adopted a wireless solution. The WLAN installed offers an 11-Mbps wireless link in a coverage area of 60,000 square feet. The technology used consists of an RF DSSS, which allows the management of large files. PCs and laptop computers can be connected to the LAN using a WLAN adapter. Besides providing Internet access, the WLAN infrastructure is also used for administrative purposes and e-mail services. Calhoun County High School: Calhoun County High School and John Ford Middle School in St. Matthews, South Carolina, needed fast, reliable, and flexible Internet access that could be used in various classrooms to capture and retain the students attention. Furthermore, the schools needed a LAN architecture that would

154


provide facilities for the mobility of teachers equipment to make equipment switching from one classroom to the next one in the timetable easy. The school had an installed wired drop in every room that would be compatible with the new LAN extension. The solution adopted by the school was a 2-Mbps WLAN that was later upgraded to a faster solution. The wireless solution allowed the connection of more PCs in a single classroom than those that could be connected using a wired system. In addition, computers that are installed in a classroom and that will be used only for two or three hours can easily be carried to other classrooms and connected to the LAN, staying in use if necessary. If they were connected to a wired LAN, portability would be very difficult. Equipment portability also offers a pedagogical benefit, because teachers can adapt the system to their personal teaching style. Students can remain at individual PCs, or dynamic cooperative learning groups can easily be formed. The WLAN performs quickly enough for the schools needs, and the multimedia laptops can satisfactorily download video files. However, the maintenance tasks of the wireless network, such as the imaging of machines, has a performance comparable to that achieved with the wired LAN architecture. ω

ω

Union Endicott schools: The Union Endicott School district is an example of how wireless links can be installed together with leased cabled lines and fiber optic lines. This district achieved a migration from a totally wired structure to a mixed wired and wireless architecture. The result is a high operatingspeed system with a substantial reduction in leasing and maintenance costs. The technical solution implemented is based on wireless bridges that operate with a DSSS radio link in the 2.4-GHz band, which does not require an FCC license. The Greenville College campus, in Greenville, Illinois, has an RF wireless network that covers classrooms, campus dorm rooms, and park benches. Students, teachers, and administrators can connect their computers to the network around the campus. The possibilities of wireless network connections have created a covered area that would be impossible to reach with traditional wired networks. In addition, the wireless system allows the college to avoid rewiring costs when users move from one place to another.


155

Another educational application of an RF wireless systems installation in the United States is that of Neodesha, Kansas, whose wireless communication system was built to give Internet access to the high school, two elementary schools, the board of education, and the library. Using a star topology, every building was connected with the high school with wireless DSSS-RF links. The high school was connected to an Internet service provider (ISP) using an ISDN line. With this configuration, the Internet services are accessible from any of the previously mentioned buildings, and the cost is minimal. For more information about these applications see [1]. ω

Health care scenarios: The access to updated information in real time is a very important objective in health care environments. Communications in such environments affect three important groups or elements: patients, hospital staff (physicians, nurses, pharmacy staff, etc.), and equipment.

ω

A personal communication device can be assigned to each patient after admission to a hospital. The signals are then collected by a networks distributed antenna system equipped with storage facilities where all diagnostic and personal data will be stored. Information on each patient is received by a central computer that transfers the data to the doctor who sends instructions to other members of the staff or requests a personal visit. Upon discharge from the hospital, data related to the illness will be given to the patient in a computerreadable format to be used in future examinations. Similar devices may be necessary for high-risk or chronic patients. In case of emergency, the portable communication device will transmit a signal automatically, or at the patients request, ask for help. Emergency data may be delivered by a personal communication device. An example of the application of wireless communication systems in hospitals follows: ω

The nursing staff of the Good Samaritan Hospital (Dayton, Ohio) works with portable computers that are connected to the hospitals information system through a WLAN. Using this system nurses can document medication administration at the bedside. The information is updated and accessed in real time. With this system the traditional multiple documentation system, which consists of written notes on paper and the entry of the information

156


into the computer system, has been avoided. The advantage of having a system where the information is updated in real time is that the information systems always contain reliable information and that users do not need to wait for the task of processing of information to finish. The cost of a wired system, including a connection to the network at every bedside, would have been prohibitive. The wireless solution has made the nurses job easier. Other applications of wireless communications systems in hospital environments deal with the information exchange between the hospitals pharmacy and the nursing staff. The medication for each patient is prepared and packed, and every packet is marked with a barcode. Nurses have a wireless portable device to scan the barcode on the medication package and the barcode that identifies each patient. The wireless portable device communicates with the pharmacy computer system using a wireless link. Thus, the computer checks the patient identity and the medication. Once this information has been verified, the computer system sends a confirmation signal to the wireless portable device, and then nurses dispense the medication to the patient. This experience has been adopted in the Veterans Administration Hospital (Southern California). For more information on these applications see [1]. ω

Law court scenarios: Wireless systems have a strong field of application in courtrooms and courthouses. Courtroom computer use makes access to reference documents, and the documents storage, easier. Electronic storage facilitates the organization and indexing of information, allowing easy access to documents (such as depositions, transcriptions, and memos). Documents can be available to plaintiffs, defendants, judges, and juries. Because law courts are usually situated in old buildings, a wireless solution for computer interconnection would be the most convenient technology.

ω

Buildings needing high security (museums) [2]: Wireless data communication systems applied to public buildings with high security requirements, such as museums, can be an interesting way of providing security control. Imagine a museum where every visitor is asked to carry a transceiver unit. This unit would continuously transmit a characteristic signal, which would, in turn, be received by the distributed antenna system. The processing of the received signals


157

would reveal the current position of every visitor. Upon leaving the public building, visitors would place their personal units into a special pocket to be deleted from the database. Further study is needed to identify the buildings needing such high security requirements and the communication services to be provided and to develop a functional entity model for such an application. Wireless systems are also suitable for other public buildings that experience high visitor traffic, such as ballparks, arenas, sports grounds, stadiums, and theaters. Wireless systems can be used to check entrance tickets, thus avoiding ticket fraud. To work in such a system, tickets need to be printed with an identification barcode. In a compatible installed wireless network, each of the entrance checking points would be equipped with a wireless unit that was, in turn, connected with a central computer containing a record of all sold tickets. Every time a ticket is scanned at an entrance checking point, the ticket data are transmitted to the central computer to validate the ticket in real time. Once the ticket has been validated, the central computer transmits entrance approval to the entrance checking point [1]. ω

Public transportation scenariosstations and airports: Airports have become very sensitive to the need to keep flights on schedule. Thus, the development of a wireless communication network to control passenger flow and the delivery of short messages within the airports is of great interest. An electronic and/or magnetic card given to the passenger upon checking in can function not only as a boarding card but also as a pager or a badge. The network under study controls passenger flow and transmits information to passengers. The broadcasting is performed by a distributed antenna system with full coverage of the airport, while the security and flow control are carried out by reading the card with special devices installed at the entrances of each of the airports rooms. In such an environment, a mixed IR and RF network can be used to transmit and display the information to the users. The benefits are security enhancement, the minimization of departure delays, and the provision of better broadcast services to passengers. The architecture of such a system must satisfy several functional constraints from both the passenger and the network operators point of view:

158


Passenger constraints: Minimum delay and user-friendly;

ω

Operator constraints: Simplification of communication protocols, easy integration, expandability, manageability, adaptability, security, and minimum human resources.

5.2.2

Business environment

The possibilities of application of wireless communication systems in office environments depend on the nature of the office, the functions carried out, and the real business purposes of the office [3]. The main advantages offered by wireless communication systems are mobility, portability, and reconfiguration requirements. It is well known that the costs of wire installation are significant and that, in some cases, wired systems are inadvisable, as in old or historic buildings. The installation of wireless communication systems in an office depends on the communication infrastructure already installed in the office. The needs may vary: ω

Extension of the existing indoor wired networks;

ω

New installation of a wireless communication system;

ω

Wireless interconnection between close buildings that want to connect their existing wired networks.

The installation process of wireless networks is faster and cheaper than the installation of wired networks. Wireless networks can be installed to connect new or temporary work groups in an office, to connect portable equipment, and, in general, to expand existing wired networks. Today, commercially available equipment for installing wireless networks has performance qualities (speed, security, reliability, etc.) comparable with the equipment for wired networks. As a result, professional staff involved in the communications infrastructure design can expect to find wireless equipment that is flexible and easy to integrate into wired systems. As an example of wireless network installation, consider the case of the Bank of America Securities LLC (BAS). BAS is a full-service investment bank and brokerage firm. Quick communication between BAS and the clients and the monitorization of changes in the market are


159

fundamental for the organization. Accordingly, it has installed a WLAN that allows workers to securely monitor the market from any place in the building. The wireless network also allows the company to set up temporary work groups in an easy and flexible way [1]. 5.2.3

Domestic buildings (the home)

Communication in the home environment deals with the exchange of control information as well as video and sound signals. Several solutions have been studied to link home equipment, each type of which uses a different technology, such as communications through the power line, wireless IR, RF, and microwave (MW) communications. Communications in a home environment include house-wide communications, connection to the public networks, entertainment services, and energy management. It is interesting to classify domestic equipment to define the different communications needs of each. Following the proposal in [4], domestic equipment can be classified as follows: ω

Brown goods (stereo, TV, VCR);

ω

White goods (washer, refrigerator);

ω

Telephone and computer communications equipment;

ω

Data-processing systems;

ω

Security/safety systems;

ω

Environment controllers.

To choose a communication technology, it is important to take into account where the equipment is placed in the house. Wireless IR links can be used for communication between equipment situated in the same room. For communication between pieces of equipment placed in different rooms, the communication technology needs to be RF or MW. Wireless links are preferred in the home environment to avoid wire installation along skirting boards and around doors and to avoid drilling holes in the walls. Also, wireless links allow changes in equipment distribution, enabling users to avoid changing network connection points, which are fixed in wired networks. An in-house wireless network provides mobility and flexibility, offering all family members access to the systems services and allowing the sharing of printers or Internet access.

160


Wireless communications systems are easy to install. They offer modularity in such a way that adding new equipment or removing old equipment is easy to do. In addition, these systems are cheaper than the traditional wired ones. It is important that wireless systems be easy to use by everybody in the house. An important factor in in-house wireless networks is the privacy of the information traveling through the network. This information can contain telephone signals, data signals from a personal computer, or video signals that cannot be accessed by unauthorized personnel. On the other hand, interference with neighboring systems must be avoided, meaning that a network installed in a house cannot interfere with the network of the neighboring house. Accordingly, wireless systems for in-house applications must include mechanisms for information protection. HomeRF is an association comprised of companies interested in this field of application. Section 5.3.1 briefly describes its activities. 5.2.4

Industrial sector

Wireless communication technologies have an important field of application in the industrial environment. Two different wireless applications can be considered in an industrial environment. One of them is open-air wireless communications, which can be interesting for those industries that have several buildings, each one with its own data network. These industries spend a lot of money on telephone lines for building interconnection. Using wireless links for these purposes would decrease telephone line costs considerably. On the other hand wireless links can be used in industry for indoor communications purposes. This allows people working in the company to be able to monitor any information involved in the production process anywhere and at any time. Using these systems, all the documents involved in the manufacturing process can be electronically formatted, avoiding the need for paper copies of documents. This means that the information related to such items as product orders, tickets, reports, and invoices can be located in a central computer. Thus, the information is available in real time and accessible from anywhere in the company. Moreover, the installation of wireless communication systems in manufacturing plants, distribution centers, or production lines makes the management of control processes easier. Reconfiguration processes are also easier due to the reconfiguration facilities inherent in wireless systems.


161

Several wireless communication systems have been installed in the industrial sector. For example, the Ford Motor Company has installed WLANs in its product plants and distribution centers. As a result of the installation, the company has noticed that the identification of recurrent quality problems takes place in real time, instead of days. This has resulted in higher productivity, cost savings on rework, and higher product quality [1]. Similarly, the Bell-Carter Olives Co. in Lafayette, California, has installed several wireless bridges for the interconnection of networks installed in different buildings to avoid the cost of traditional phone lines [1]. As in other environments, the wireless communication medium in the industrial sector can be IR (LED, laser), MW, or RF. The choice must be made according to environmental factors, such as the presence of obstacles, thermal radiation, range, and EMI.

5.3

Wireless LAN technologies and products

In the last few years a great deal of research directed at developing wireless communications systems has taken place. The research activities have facilitated the knowledge of the channel transfer function of indoor environments, as well as the development of wireless transmitter and receiver devices for both technologies: RF and IR. Electrical circuits for wireless signal processing have been developed, and the design of communication protocols for information exchange using wireless links has yielded commercial equipment ready for use. Several companies are involved in the development and commercialization of wireless products. Many of them have joined common organizations. In general, such organizations fall into one of two categories depending on whether their work is based on RF or IR technology. 5.3.1

The RF market

There are several organizations involved in the research, development, and commercialization of wireless RF communications systems and products: ω

WLANA: A trade association that does not focus on a single area but deals with the technology areas in use at each moment;

162


WECA: A technology alliance that is devoted, broadly speaking, to promoting the development of the Wi-Fi (802.11 HR) technology;

ω

Bluetooth SIG: A technology alliance devoted to promoting the development of Bluetooth technology;

ω

WLIF: A technology alliance devoted to promoting the OpenAir 802.11 FH technology;

ω

Home RF: A technology alliance that promotes home RF technologies (SWAP-CA, SWAP-lite, and SWAP-MM);

ω

The Broadband Wireless Internet Forum (BWIF): A program of the IEEE Industry Standards and Technology Organization (IEEE-ISTO) that consists in an incorporated not-for-profit association of industry-leading companies.

The following subsections briefly review these organizations. WLAN Association (WLANA) The WLAN Associations Web site is very well-organized and contains a lot of interesting information. We strongly recommend that readers visit the site at http://www.wlana.com. As is stated on the groups home page, WLANA is a nonprofit consortium of WLAN vendors established to help educate the marketplace about WLANs and their uses. WLANA develops educational materials on WLAN users experiences, applications, and industry trends. The associations Web site includes industry studies, white papers, application stories, and links to related topics and member Web sites. Table 5.1 lists WLANAs current sponsor companies. Table 5.1 WLANA Sponsor Companies Company

Web site

3Com Corporation

http://www.3com.com

Aironet

http://www.aironet.com

Breezecom

http://www.breezecom.com

Enterasys Networks

http://www.enterasys.com

NoWiresNeeded

http://www.nowiresneeded.com

Intersil

http://www.intersil.com

Intermec

http://www.intermec.com

Symbol

http://www.symbol.com


163

WLANA offers its affiliate member program to those companies that are interested in the application of WLANs but do not manufacture WLAN products. Table 5.2 lists the current WLANA affiliate members. Following its plan for the promotion of educational activities on the applications, trends, technologies, and other topics related to WLANs, WLANA presents a brief introduction to WLANs. The topics presented include an explanation of what wireless communication is and how it can be applied to connecting equipment through a LAN or other communication structures. The presentation also includes a description of the technologies that can be used to install a wireless communication system and how these systems can be configured depending on the application. As examples of successful installed systems, the Web site includes several cases classified by the company that supplies the equipment and by the type of application. Cisco Systems (Aironet), BreezeCom, Intermec, Nokia, No Wires Needed, and Symbol are the companies that stand out by their presence in the wireless connectivity world. The list of applications can be found on the following: ω

http://www.aironet.com/markets/, which contains the activities from Cisco;

ω

http://www.breezecom.com, which contains the activities from BreezeCom;

ω

http://www.intermec.com, which contains the activities from Intermec;

ω

http://www.nokia.com, which contains the activities from Nokia;

Table 5.2 WLANA Affiliate Members Company

Web site

WRQ

http://www.wrq.com

The Phillips Group

http://www.thephillipsgroup.net

Broad IP Sybergen Networks

http://www.sybergen.com

Mind Matrix

http://www.mindmatrix.net

The Walt Disney Company http://www.disney.go.com

164


http://www.nowiresneeded.com/, which contains the activities from No Wires Needed;

ω

http://www.symbol.com/products/wireless/wireless_business_solut.html, which contains the activities from Symbol.

Focusing on the markets in which wireless systems are increasing their presence we find applications in the following fields: ω

Health care;

ω

Trading and banking;

ω

Sales, consulting, banking, and finance;

ω

Restaurant and retail;

ω

Warehouse and manufacturing;

ω

Education and research.

Important information included in the WLANA Web site is the information contained in the resource directory, where the visitor will find several documents about the benefits of installing a WLAN system or about the state of the WLAN industry. Suggested readings and related Web sites are also included. The news room includes articles on recent industry events, industry magazines, and press releases. Finally, the Web site includes an interesting forum where visitors can present their questions, answers, opinions, and related information. Wireless Ethernet Compatibility Alliance (WECA) WECA is useful for the users of wireless equipment because WECA-certified products have passed an interoperability test on WLAN products. WECA is defined as the organization created to certify the interoperability of Wi-Fi TM (IEEE 802.11 high rate) WLAN products and to promote Wi-Fi TM as the standard for WLAN deployments across all market segments. The current sponsors of this organization are listed as follows:

2Wire 3Com Alantro Communications AMD

Apple Artem Askey Atheros


Atmel Breeze Com Buffalo Cisco Systems Compaq Dell Elsa Emtac Enterasys Eumitcomm Fujitsu Gemtec HWN IBM Intel Intermec Intersil Industrial Technology Research Institute of Taiwan

165

Lucent Technologies MobileStar Networks Mobilian Nokia No Wires Needed Phillips Samsung ShareWave Siemens Sony Spectralink Symbol Teklogix Telxon Toshiba Wayport Wireless Solutions Xircom Z-Com Zoom

If a company would like to submit a product to the WECA certification laboratory for interoperability testing and certification, the company must be a member of WECA. Membership requirements and benefits can be found at the WECA Web site. When a product meets its interoperability requirements, WECA grants a certification of interoperability, which allows the vendor to use the Wi-FiTM logo on advertising and packaging for the certified product. The Wi-Fi seal of approval assures the end users of interoperability with other network cards and APs that also bear the Wi-Fi logo. WLAN Interoperability Forum (WLIF) When literally transcribing the WLIF mission from the associations Web site, it can be found that the WLIF was formed to promote the use of WLANs through the delivery of interoperable products and services at all levels of the value chain. Members of WLIF believe that open competition between compatible products will benefit customers by assuring them that their WLAN will support the best products on the market today and offer unparalleled flexibility and adaptability for the future.

166


The aim of this forum is to provide a review of all open-air and IEEE 802.11 products. Products certified by WLIF include APs, adapters, pen computers, scan/key handheld products, serial modems, handheld printers, PDAs, and vehicle mount units. With the information provided by WLIF, the wireless system designers are free to choose the products from any manufacturer that best meets their requirements. It is supposed that in such a competitive environment, suppliers will improve their products, reduce prices, and offer a better service. Other information that can be found at the WLIF Web site includes technical white papers, a list of WLIF members, a certified products listing, and news and press releases. Information on the Bluetooth SIG organization can be found at http://www.bluetooth.com/. The companies that lead this organization are 3Com, Ericsson, Intel, IBM, Lucent, Microsoft, Motorola, Nokia, and Toshiba. As specified on the Web site, the Bluetooth SIG is a forum for enhancing the Bluetooth specification and providing a vehicle for interoperability testing. More than 1,200 companies have joined the Bluetooth SIG. The advantage of Bluetooth SIG is that the resulting organization is greater than the sum of its parts. Product solutions will be the result of a concerted effort between the involved companies. The Web site includes interesting information, such as the release of the Bluetooth 1.0 Specification. In addition, there is an information section devoted to Bluetooth products, including components, consumer products, prototypes, and development tools. The organization also provides an explanation about what the Bluetooth technology is and how it can be used. Finally, as is commonly provided on Internet sites, a forum welcomes visitors opinions and questions. Bluetooth Special Interest Group (Bluetooth SIG)

The HomeRF Association The information about the HomeRF Association can be found on http://homerf.org. According to the Web site, the mission of the HomeRF Working Group is to enable the existence of a broad range of interoperable consumer devices, by establishing an open industry specification for wireless digital communication between PCs and consumer electronic devices anywhere in and around the home. The HomeRF Association deals with the following consumeroriented applications of wireless RF communications systems, all of which are discussed on the associations Web site:


167

ω

RF digital communications for PCs and consumer devices anywhere in and around the home;

ω

Wireless Internet access and printer sharing in and around the home;

ω

Multiple PCs sharing a single Internet connection;

ω

The sharing of files, drives, and printers without wires;

ω

The reception of MP3 music;

ω

Cordless data and voice, taking full advantage of a broadband Internet connection.

The information available on the Web site includes the necessary products to install a wireless world at home. The HomeRF Association has a working group. The activities of this working group are the development of a specification for wireless communications in the home. With this purpose, the group has developed the shared wireless access protocol (SWAP). The technical summary of the SWAP specifications is available at the HomeRF Web site. The HomeRF alliance is made up of members that have joined the alliance as adopter members, participant members, or both. Membership information and benefits are available at the Web site. Other interesting information to be found there are FAQs, an events calendar, and HomeRF news stories. Interested readers should visit the HomeRF Web site. Information about this forum can be found on http://www.bwif.org. According to this Web site, members of BWIF are committed to drive product road maps that will lower product costs, simplify deployment of advanced services, and ensure the availability of interoperable solutions based on VOFDM technology. BWIF members agree to cross-license to other BWIF members the technologies required to implement the VOFDM specifications on a worldwide, royalty-free basis. The goal of BWIF is to facilitate cost-effective, broadband wireless access solutions, with industry leading performance and reliability for compelling end-user applications such as high-speed Internet access, premium streaming audio and video content, and voice. BWIFs current promoter members are: Broadcom Corporation, California Amplifier, Cisco Systems, Pace Micro Technology, PipingHot The Broadband Wireless Internet Forum (BWIF)

168


Networks, Texas Instruments, Toshiba, TurboNet Communications, and WJ Communications. The organizations adopter members are Adicom Wireless, Bechtel, Fluor, Getronics, LCC International, Moseley, nBand Communications, Oren Semiconductor, Remec, RF Solutions, Telaxis Communications, WaveIP Ltd., WFI (Wireless Facilities, Inc.), and ZyGATE Communications Inc. This forum was created only recently, so we recommend that interested readers visit the Web site to obtain updated information. Other associations available on the Internet are listed as follows: ω

Hiperlan 1 Global Alliance (H1GA) at http://www.hiperlan.org/. This is a technology alliance to promote HiperLAN1 interoperability.

ω

Hiperlan 2 Global Alliance (H2GA) at http://www.hiperlan2.com/. This is a technology alliance to promote HiperLAN2 interoperability.

ω

IEEE 802.11 at http://standards.ieee.org/, whose activity deals with the written specification of the standard body.

ω

IEEE 802.15 at http://standards.ieee.org/, whose activity deals with the written specification of the standard body.

ω

ETSI HiperLAN at http://www.etsi.org/, whose activity deals with the written specification of the standard body.

ω

ETSI BRAN at http://www.etsi.org/, whose activity deals with the written specification of the standard body.

ω

University of New HampshireInteroperability Lab (UNH-IOL) at http://www.iol.unh.edu, which is an independent interoperability test laboratory.

5.3.2

The IR market

Once the communications industry had identified the applications of IR wireless communication systems, several devices, equipment, and systems were launched on the market. IR wireless communication systems are continuously evolving, and the connectivity possibilities offered by these systems are driven by the development of new products. Several companies interested in wireless IR systems for data communications have joined together to form IrDA, the aim of which is to create and promote interoperable, low-cost IR data interconnection standards that support walk-up, point-to-point user models. The standards


169

developed by IrDA can be used in many data communications applications, such as communications between PCs and peripheral devices, digital image exchange, communications between mobile devices, modem access, and exchange between MP3 players and computers. Bit rates using IrDA standards started with 9,600 bps to 115 Kbps. Later specifications deal with data rates of 2 Mbps, 4 Mbps, and 16 Mbps. Appendix 5A lists the companies that comprise IrDAs 2000 membership. IrDAs Web site, http://www.irda.org, is worth visiting. It contains a design guide for technical professionals who want to design IR wireless communications systems or to integrate IrDA products. The design guide includes several topics: ω

Hardware guides, with information and tools to assist during system design and the integration process;

ω

Software guides, with information and tools to design applications that use IR wireless communications;

ω

Component selection guides, with the information necessary for selecting IR wireless components;

ω

Interoperability testing, including test plans and processes as well as a listing of independent test labs;

ω

Information on the comparison of IrDA technology with other wireless technologies;

ω

IrDA references, including articles, books, presentations, and papers;

ω

Information on tutorials, courses, and conferences.

IrDAs Web site also includes a brief explanation on how IrDA systems work and information on how products can work together. A FAQ section is also available as well as a public news group where users can share information on these systems. Table 5.3 details IrDA components manufactured by IrDA companies as well as their activities. As Table 5.3 shows, there is a great deal of activity in component manufacture, with 19 companies offering IR emitter devices, photodiodes, and chips that can be used to develop the physical layer of IR wireless communications systems. There are also 13 companies offering transceivers suitable for electrical-to-optical signal conversion and

170


Table 5.3 Components Manufacturers

Company

Chips

Photodiodes

LEDs

I/O

Transceiver receiver encoder decoder

ACTiSYS

√

√

Boards

Hardware cores

AMD

√

Calibre

√

√

√

√

Califonia Eastern Labs

√

Dowa

√

√

Hewlett-Packard

√

√

√

√

Infineon Technologies

√

Intel

√

Motorola

√

National Semiconductors

√

NEC

√

Novalogic

√

√

√

Okaya Systemware

√

Panasonic

√

√

√

Parrallax

√

√

Phoenix Technologies

√

Rohm

√

√

√

√

√

Sanyo

√

Sharp

√

√

√

Stanley Electric

√

√

√

√

√

TEMIC

√

√

√

√

√

Tekram

√

√

Texas Instruments

√

√

√

√

√

Toshiba

√

Unity Microeletronics

√

Winbond

√

√

√


171

encoders and decoders for signal processing. Other products such as boards and hardware cores are supplied by nine different companies. Table 5.4 details IrDA companies that offer mobile computing products. As Table 5.4 shows, most of the mobile computing products are related to notebooks, portables, and handheld PDAs. Although these pieces of equipment are not connected to a communication system during their normal operating mode, they have to communicate with fixed equipment, such as desktop PCs or printers for such functions as Table 5.4 Mobile Computing Products

Company

Notebooks/ portables

3Com

Handheld PDAs

Adapters/ dongles

Digital electronic capture devices

Print- Teleers phony

Network access

Cameras

√

√

√

√

√

√

√

Acer Laboratories

√

ACTiSYS

√

AMD

√

AMP

√

Apple

√

√

Canon

√

√

Casio

√

√

Citizen America

√

Clarinet Systems

√

Compaq

√

√

Dell

√

Eastman Kodak

√

Ericsson

√

Extended Systems/ Counterpoint

√

√

Fujitsu

√

√

172


information exchange, backup, and printing functions. It is better to have a wireless link for these purposes. That is why most of them include an IrDA port. Table 5.5 details the IrDA companies that manufacture products for desktop computers and peripheral devices. Table 5.5 shows how wireless IR communications ports are integrated in fixed equipment. As previously stated, this is an interesting

Table 5.5 Products for Desktop Computers and Peripheral Devices Company

PC

Printer

Adapter/ dongle

Telephone

Modem

Keyboard

3Com

√

√

√

√

√

√

√

√

3Jtech

√

√

ACTiSYS

√

√

AMP

√

Canon

√

CGP electronics

√

Extended System/ Counterpoint

√

FCC Limited

√

HewlettPackard

√

√

Mouse

Remote control

Matsushita

√

Megatec

√

√

NEC

√

√

Nokia

√

Novalog

√

ONeil

√

Okaya Systemware

√

√

Parallax

√

√

REUDO

√

Sharp

√

Tekram

√

√

Toshiba

√


173

communication system for data transfer from portable equipment (such as notebooks or PDAs). It is important to note that the majority of manufacturers include wireless IR ports in their equipment. Table 5.6 details the IrDA companies that are software providers. Software tools are important in managing the hardware that allows IR wireless communications to take place. These software providers offer a wide range of programs for managing IrDA ports that are easy to include in an operating system. Notice that the two most important PC operating system suppliers (Microsoft and Linux) are involved in the development of tools for wireless IR communications. Table 5.6 Software Products

Company

OS/OS extensions/ protocol stacks/ drivers

Development tools

Consultant services

Application programming

3Com

√

√

√

√

Access

√

ACTiSYS

√

√

√

Calibre

√

Eastman-Kodak

√

EMBEDnet Inc.

√

Extended Systems/Counterpoint

√

√

√

√

Extended Systems

√

FCC Limited

√

Geoworks

√

HewlettPackard

√

Linux

√

Microsoft

√

Microware

√

Motorola

√

√

Norand

√

√

√

√

Okaya Systems

√

√

Open Interface

√

√

√

174


Table 5.6 (continued)

Company

OS/OS extensions/ protocol stacks/ drivers

Development tools

Consultant services

Application programming

Parallax

√

√

√

√

Phoenix

√

Puma

√

√

Questra

√

√

REUDO

√

√

TEMIC

√

Trace Research Center

√

Finally, Table 5.7 lists the IrDA companies that are service providers. The services they offer are testing, integration, and consultancy. Hardware and software tests assure the proper operation mode of equipment. In addition, IrDA has developed a standard interoperability test. Products passing the IrDA compliance and IrDA interoperability tests are awarded the status of IrDA reference products and can show the IrReady logo. An independent test lab carries out IrDA interoperability tests. Many companies offer hardware and software consultancy and integration services, as shown in Table 5.7. There are applications for IR wireless systems other than data communications applications. The remainder of this section presents the activities of three companies that have devoted a great deal of time and effort to developing IR wireless applications during the last few years. First, consider ELPAS Smart Wireless Environments. Its main product line is the Local Position Systems (LPS) and the application of EIRIS LPSs in smart hospitals and intelligent buildings. EIRIS Healthcare consists of a network of readers (RDRs) mounted on the ceilings of rooms that communicate with mobile personnel, patients, and equipment badges, using wireless IR signals that are eyesafe and do not interfere with other electronic equipment. RDRs are connected to a server where the information is processed. Using a PC connected to the network, any operator can see the information in real


175

Table 5.7 Service Providers

Company

Hardware testing

Software testing

Product integration

Hardware consultant

Software consultant

Integration services

3Com

√

√

√

√

√

√

ACTiSYS

√

√

√

√

√

√

Calibre

√

√

EMBEDnet Inc.

√

Extended Systems/ Counterpoint

√

√

√

Helmig

√

Microware

√

Norand

√

√

Open Interface

√

Parallax

√

√

√

Phoenix

√

Questra

√

√

√

Waseda University Test Lab.

√

time. This allows for the identification and location of people and equipment. Mobile equipment can also send alert signals. EIRIS intelligent buildings is a similar application with the same network architecture based on RDRs. The users of this system enjoy automated entrances, individualized electronic signage, automatic lift calling, identification and location of equipment in real time, individualized lighting and temperature control, and personalized hands-free computer access. Finally, the ELPAS AIRnet provides an Ethernet wireless IR LAN connection. The architecture of the wireless network consists of several ceiling-mounted units that interface with the wired LAN. Any equipment

176


having an Ethernet port can be connected to the wireless network using a portable unit that allows direct communication with other equipment (peer-to-peer) or communications with the network through the ceiling unit. For more information visit the companys Web site at www.elpas.co.il. JOLT is also an important company in wireless optical communications. It offers wireless communications systems to connect buildings using point-to-point links. Its wireless equipment for data transmission abides by the following standards: ATM-155-Mbps, FDDI, fast Ethernet, token ring, and Ethernet. The company also offers wireless connectivity for telephony applications (E1/T1 to E3/T3 lines, OC3), cellular systems (with microcell and picocell connectivity), video transmission (multimedia and surveillance), and intra-office connectivity for open space work areas. JOLT systems are suitable for networking in urban environments and campuses, for crossing difficult terrain (such as highways, rivers, and airports), or for temporary installations. For more information visit the companys Web site at www.jolt.co.il. Oplink Communications is another company whose aim is the development and manufacturing of wireless optical communications products. It offers several wireless IR communications systems that can be used in different environmentsboth indoor and outdoor. These systems allow mobile data transfer, and they can be used for fleet management. Other fields of application are transport and security, audience response systems, wireless barcode reader communications, medical monitoring, data collection (with fixed and mobile terminals), wireless IR computer networking, agricultural applications, rotating systems connectivity (optical slip rings), warehouse management, military applications, ISDN wireless extenders, and industrial applications. Their main markets are office, industrial, transport, and medical. For more information on system specifications (i.e., data rates, maximum range, and protocols) visit the companys Web site at www.oplink.com.

5.4

Conclusions

Most of the leading companies in the area of information technologies have a product line related to wireless communication systems. This is significant in evaluating the importance of wireless technology. The possibilities offered by this technology allow a decrease in the installation costs of the wired communication infrastructure. From the system users


177

point of view, wireless systems are easier to use and handier than wired systems, because all of the plugging and unplugging processes are avoided. Wireless technology also allows users to undertake projects and implement ideas that were impossible to achieve using wired systems. Applications such as local positioning systems that help users find people in a hospital or a company are possible thanks to wireless systems. Today, data rates through wireless communications systems and their QoS are comparable to those of wired systems. Therefore, these parameters are no longer a constraint. Accordingly, it is reasonable to expect a boom in the growth of installations including wireless systems.

References [1]

www.aironet.com/markets.

[2]

Sarantopoulos, S., Public Building Environment in Wireless In-house Network Studies, European Commission, Luxembourg, 1994.

[3]

Takahashi, O., and T. Touge, Optical Wireless Network for Office Communications," JARECT, Vol. 20, Telecommunication Technologies, 1996, pp. 217228.

[4]

Marinakis, D., and A. Najand, User Requirements in the Home Environment in Wireless In-House Network Studies, European Commission, Luxembourg, 1994.

Selected Bibliography Barry, J., Wireless Infrared Communications, Kluwer Academic Publishers, 1994. Crow, B., et al.; IEEE 802.11 Wireless Local Area Networks, IEEE Communications Magazine, Sep. 1997, pp. 116126. Elliot, S., and D. Dailey, Wireless Communications for Intelligent Transportation Systems, Norwood, MA: Artech House, 1995. Pahlavan, K., and A. Levesque, Wireless Information Networks, New York: J. Wiley & Sons, Inc., 1995.

178


Appendix 5A: IrDA membership Company

Web site

3COM/Palm Computing

http://www.palm.com

Access

http://www.access.co.jp/english

Acer Laboratories

http://www.acer.com

ACTiSYS

http://www.actisys.com/IrDAProd.html

Agilent Technologies

http://www.semiconductor.agilent.com/ir/index.html

A. I. Corporation

http://www.aicp.co.jp

AMP

http://www.ampincorporated.com

AMSKAN

http://www.amskan.com

Anritsu

http://www.anritsu.com

Apple Computer

http://www.apple.com/products

Assn Interactive Media

http://www.interactivehq.org

Avio Digital

http://www.aviodigital.com

Calibre

http://www.calibre-inc.com

California Eastern Labs

http://www.cel.com

CANAL+

http://www.cplus.fr

Canon

http://www.canon.com

Canon Systems Globalization

http://www.usa.ccsi.canon.com

Capella Microsystems Casio Computer

http://www.casio.co.jp

Citizen Electronics

http://www.c-e.co.jp/english

CITIZEN WATCH

http://www.citizenwatch.com

Clarinet Systems

http://www.clarinetsys.com

CMD Technologies

http://www.cmd.com

Compaq Computer

http://www.compaq.com


http://www.countersys.com

Credicom Technologies

http://www.credicom.com

CrossCheck

http://www.cross-check.com

Data Dimensions

http://www.data-dimensions.com

Datalogic

http://www.datalogicsys.com

Dell Computer

http://www.dell.com

DIGI-TEK Laboratory

http://www.digi-tek.com

Digital Print

http://www.digitalprint.com

DOWA

http://www.dowa.co.jp/english/semicon/index_e.htm

Eastman Kodak

http://www.kodak.com

EMBEDnet Inc.

http://www.embednet.com


Company

179

Web site

Ericsson

http://www.ericsson.com


http://www.extendsys.com/products/infrared

First Web Bank

http://www.1stwbd.com

Fuji Photo Film

http://home.fujifilm.com

Fuji Xerox

http://www.wired.com/news/news/story/3949.html

Fujitsu

http://www.fujitsu.com

Funai Electric

http://www.funai.co.jp/funai/english/eg_index.html

FusionONE

http://www.fusionOne.com

Glenayre

http://www.glenayre.com/main.asp

Helmig Engineering

http://ourworld.compuserve.com/homepages/helmig_eng

Hewlett-Packard

http://www.hp.com/go/ir

Hitachi

http://www.hitachi.com

Holtek

http://www.holtek.com.tw

I.D. Factory IBM

http://www.ibm.com/products

Infineon Technologies

http://www.smi.siemens.com/opto

InFocus

http://www.infocus.com

Integrated Systems Intel

http://www.intel.com

IR DATA Corporation

http://www.irdatacorp.com

ITE

http://www.ite.com.tw

Inventec

http://www.inventec.com

JVC

http://www.jvc-victor.co.jp

Kawasaki Steel

http://www.kawasaki-steel.co.jp/index_e.html

KC Technology

http://www.kcti.com

Kent Ridge Digital Labs

http://www.krdl.org.sg

Kobe Steel

http://www.kobelco.co.jp/indexe.htm

Kodenshi Korea

http://www.kodenshi.com

Linux-IrDA Project

http://www.bratti.net/dag

Logitech

http://www.logitech.com

Matsushita ElectronicWorks

http://www.maco.panasonic.co.jp

Matsushita/Panasonic

http://www.panasonic.co.jp/global/index.html

Maxim Integrated Products

http://www.maxim-ic.com

Microsoft

http://www.microsoft.com/hwdev/infrared

Minciu Sodas Laboratory

http://www.ms.lt

Minolta

http://www.minoltausa.com

180


Company

Web site

Mitsubishi

http://www.mitsubishi.com

MKNet Corporation

http://www.mknetcorp.com

Motorola

http://www.mot.com

National Semiconductor

http://www.national.com

NEC

http://www.ic.nec.co.jp/opto/index.html

NetSchools

http://www.netschools.net

New Japan Radio

http://www.njr.co.jp/index_e.htm

Niigata Canotec

http://www.niigata-canotec.co.jp/welcome.html

Nokia Mobile Phones

http://www.nokia.com

Nova Crystals

http://www.novacrystals.com

Novalog

http://www.novalog.com

NTT DoCoMo

http://www.nttdocomo.co.jp

NTT/Nippon Tel & Tel

http://www.ntt.co.jp

Okaya Systemware

http://www.osw.co.jp

Open Interface

http://www.openinterface.com

Extended Systems/Parallax

http://www.parallax-research.com

PDAia Pentax Systems

http://www.pentax.com

Personal Solutions Philips

http://www.philips.com

Phoenix Technologies

http://www.phoenix.com

Pumatech

http://www.pumatech.com

REUDO

http://www.reudo.co.jp

Ricoh

http://www.ricoh.com

ROHM

http://www.rohm.co.jp

Ryoyo Electro Co.

http://www.nkk.co.jp/LSI/sales/sales.html#Sales.US

Salutation Consortium

http://www.salutation.org

Sanyo

http://www.sanyo.com

Scientific Atlanta

http://www.sciatl.com/nav/html/top/loframe.html

Seiko Instruments

http://www.sii.co.jp

Sejin

http://www.sejin.com

Sharp Electronics

http://www.sharp-usa.com

Sigmatel

http://www.sigmatel.com/products.htm

Silitek

http://www.silitek.com

SMK Manufacturing

http://www.smkusa.com

Sony

http://www.sony.com

Stanley Electric

http://www.stanleyelec.com


Company

181

Web site

SteelCase

http://www.steelcase.com

Strategies Unlimited

http://www.strategies-u.com

Sun Power

http://www.sunpower.com

Tekram Technology

http://www.tekram.com

Texas Instruments

http://www.ti.com/sc/irda

Toshiba AITC

http://www.csd.toshiba.com

Trombley Takenaka International

http://www.ttijapanese.com

Unity Opto Technology Universal Electronics

http://www.ueinc.com

USB Visa International

http://www.visa.com

Vishay-Telefunken

http://www.vishay.de

Waseda University Test Lab Winbond

http://www.winbond.com.tw

ZiLOG

http://www.zilog.com

.

CHAPTER

6

Contents 6.1 Introduction: Is the future wireless?

Upcoming Standards and Future Trends

6.2 The evolution of HIPERLAN

J. M. Riera and R. Peréz-Jiménez

6.3 The evolution of IEEE 802.11 6.4 Forthcoming IR standards 6.5 Other RF standards: DECT, Bluetooth, WATM, HomeRF, etc. 6.6

Conclusions

6.1 Introduction: Is the future wireless? In the last decade, a large number of standardization bodies, official or industrial, have been busy working on the definition of wireless (RF and IR) systems for applications ranging from high-speed kilometer-range wireless access (i.e., local multipoint distribution system (LMDS)) to low-speed veryshort-range device interconnection (i.e., the original IrDA). When considering the forces that drive this work the world over, some important points must be taken into account: ω

In the alternative between standards and proprietary systems, the former seems to be winning in all moderate- to large-size markets because of the proven effectiveness of economies of scale in reducing manufacturing costs. Besides, users are 183

184


demanding interoperability to be free to choose equipment from different manufacturers. The global successes of standards such as GSM for mobile communications or Ethernet in the LAN market are pumping the standardization works. ω

The official standardization bodies are seen as slow by some parties, due to the fact that they are subject to pressures from different agents. Some industrial corporations have formed alliances to define specifications intended to become industrial de facto standards. IrDA or Bluetooth are good examples of this. This does not exclude the fact that some industrial standards may become official standards in the future.

ω

The information society is bringing about huge interchanges of data that must be carried to the user in very different situations for a wide range of applications. The flexibility provided by wireless links can be a determinant in the rapid growth of many applications.

ω

The success of mobile communications, with mobile phones approaching or surpassing the number of fixed phones in many countries, will almost certainly be followed by the success of mobile data communications.

ω

Cabled communications involve the need for infrastructure that is hard to install, and can be very inconvenient. For example, the cost of cabling cities is high, and the process takes many years. In domestic/business premises, the cost of the cables is not large, but the user is fixed to the end of the cable, and there is some burden in the need to have cables all around. A computer without the keyboard, mouse, and printer cables (which is technically feasible today) is not only much cleaner and more elegant, it is also easier to install and transport from one place to another and more comfortable to use as the different elements are not mechanically linked.

However, cabled networks are usually cheaper, which is largely due to economies of scale. They can also provide very high bandwidths, in excess of what can be reasonably provided by wireless networks. New applications such as the family of xDSL [1] or cable modem are the keys to the future of cable communications in access networks. A wireless future can be envisioned with the following features:


185

ω

Telephony will be mostly mobile, with phone numbers associated with people, instead of places. The number of mobile phones will be on the order of 90% of the population able to use it (everybody over six, with the possible exception of disabled or very old people).

ω

Fixed access, which will still be necessary for the home and offices, will be provided by microwave and millimeter-wave links, with point-to-multipoint standards (like LMDs), or, in the case of high bandwidth users, point-to-point links.

ω

Inside the home or office, devices will communicate to each other with some of the short-range technologies (WLAN or others) that are currently being defined.

ω

Internet access, or data access in general, will be provided at fixed points by the mentioned fixed access technologies, and by third and future generations of mobile communications. Very-highspeed broadcast can be obtained from satellites (with DVB transmission) and future stratospheric platforms. Entertainment, such as TV, will eventually become indistinguishable from Internet access.

This wireless future is a maybe, as cable communications remain very active, and the scarcity of the radio spectrum will always remain a hard limit to the capacity that can be carried. However, many of these features are present today or will arise in the near future. This book focuses on WLAN standards. The rest of this chapter discusses the evolution of the wireless standards discussed in previous chapters (IEEE 802.11, HIPERLAN, and IrDA), and presents new wireless standards that can provide LAN services but that have been defined for a broader range of applications. This is indicative of a developing trend: standards that cover multiple applications, providing a link to communication terminals while offering enough flexibility to be employed in a variety of scenarios. Another important trend that is worth mentioning is the evolutionary nature of current standards. They are not fixed once and for all but are being updated again and again to add new features or to find new applications. The current speed of technological innovation is determinant on this fact. On many occasions, the new version of a standard can

186


be implemented with the same hardware, by means of updating the software.

6.2

The evolution of HIPERLAN

Chapter 4 extensively described the HIPERLAN/1 standard and introduced HIPERLAN/2 and the other ETSI BRAN standards (HIPERACCESS and HIPERLINK). This chapter discusses only the current state of HIPERLAN/1 and HIPERLAN/2, as HIPERACCESS and HIPERLINK are not WLAN, although they may complement one. The HIPERLAN/1 specifications are finished, and the HIPERLAN/2 specifications are currently quite advanced. HIPERACCESS and HIPERLINK specifications are still in the early phase. Some manufacturers, with the purpose of accelerating the adoption of HIPERLAN/1-based products worldwide, have formed the HIPERLAN Alliance. Market promotion, spectrum lobbying, and cooperation with other standardization bodies are among the objectives of the HIPERLAN Alliance. More information can be found in [2]. To our knowledge, there are no HIPERLAN products on the market at the moment, but they are expected to appear very soon. HIPERLAN/2, originally based on WATM, has extended its scope for providing WLAN services and for interacting with core networks of different kinds, such as ATM, IP, and UMTS. Figure 6.1 shows the reference model for HIPERLAN/2. A convergence layer provides connectivity with the core network. Thus, the standard is open in that it can use other networks in the future, with the only specification being that of the corresponding convergence layer. A good description of this standard can be found in [3], a document that also includes references to the related ETSI technical documents. The most important distinguishing characteristic of HIPERLAN/1 is its centralized MAC, which supports QoS functions. HIPERLAN/2 can operate in a centralized mode, which allows the implementation of radio access to a fixed point. Thus, a network can be implemented to cover entire buildings through the deployment of APs in a cellular structure. It can also operate in ad hoc networks, without the need for APs, but in this case one of the devices must operate as a central controller. The central controller can be any of the devices in the network, and can change if, for example, the current central controller leaves the net.


187

Radio access network Radio Terminal Convergence subsystem layers entities Terminal

Access network

Air interface

Convergence layer 1

TCP/IP

Convergence layer 2

ATM/ B-ISDN

Convergence layer 3

UMTS

Convergence layer n

IEEE 1394

Broadband radio access network Figure 6.1

Core networks

Reference model for HIPERLAN/2.

As the MAC is centralized, the radio access is structured in TDMA/ TDD frames governed by the AP or the central controller (CC). The frame may include bursts of the following types: ω

Broadcast control channel (BCCH): Network information that is distributed to all users from the AP or CC.

ω

Down-link (D-link): Information (control and data) that is sent from the central node to the others. It can also include broadcast data in some cases.

ω

Up-link (U-link): Information from the mobile terminals to the central node. The U-link resources must be requested and provided by the central node.

ω

Direct link (Di-link): Information that is directly sent from one terminal to another. The central node is also involved in the assignment of resources in this part of the frame.

ω

Random access channel (RACH): The mobile terminals can access this channel without previous assignment. It is used, for example, for first-time registrations on the net.

HIPERLAN/2 employs 20-MHz radio channels, with orthogonal frequency division multiplex (OFDM). In this multicarrier modulation, the available bandwidth is distributed between a large number of

188


Table 6.1 OFDM Parameters in HIPERLAN Type 2 Parameter

Values

Channel

20-MHz

Number of subcarriers

52 (48 + 4 pilots)

Symbol interval

4 µs

Guard interval

0.8 µs

Subcarriers separation

312.5 KHz

Total bandwidth

16.56 MHz

Subcarriers modulation

BPSK, QPSK, 16-QAM, 64-QAM

Coded bit rate

12, 24, 48, 72 Mbps

Coding rate

1/2, 9/16, 2/3, 3/4

Data bit rate

6, 9, 12, 18, 27, 36, 54 Mbps

subchannels. Each subcarrier is modulated with a phase amplitude modulation format. Because the subcarrier separation is chosen to be the inverse of the symbol interval, the subchannels are orthogonal to each other. Actually, they are frequency-overlapped, but data can be extracted without intercarrier interference if orthogonality is maintained in the transmission. Table 6.1 shows the main parameters of the OFDM in HIPERLAN/2. Data rates can be as high as 54 Mbps. Data is convolutionally encoded, with code rates of 1/2, 9/16, 2/3, and 3/4. The standardization process of HIPERLAN/2 is very advanced. A global forum has been established by some of the major manufacturers [4]. The HIPERLAN/2 standard is coordinated with new IEEE 802.11 definitions and the Japanese MMAC project [5]. More information on HIPERLAN/2 can be found in [6, 7].

6.3

The evolution of IEEE 802.11

Soon after the IEEE 802.11 specifications were closed, new PHY modes were defined to extend the standard so that data rates could be augmented from the initial 1 and 2 Mbps. The MAC is kept unchanged in the extensions. Two extensions have been defined, one to achieve 5.5 Mbps and 11 Mbps in the 2.4-GHz ISM band, and a second extension to achieve much larger bit rates in the 5-GHz band, which has been opened in the United


189

States for unlicensed use through the category of unlicensed national information infrastructure (UNII). This is roughly the same band reserved in Europe for HIPERLAN. To transmit at 5.5 Mbps and 11 Mbps in the 2.4-GHz band, a new modulation scheme is defined, based on the DSSS PHY standard. In the original PHY, a chip rate of 11 Mc/s/s was selected, with a symbol rate of 1 Msps. The data rates of 1 and 2 Mbps are obtained through the use of BPSK and QPSK. (See Chapter 3.) The chip rate is maintained in the extension to higher data rates, as is the QPSK modulation. Accordingly, much of the hardware and the channel structure are compatible with the lower data rates. However, the original DSSS modulation is no longer useful, as the chip rate is equal to or double the bit rate. The process gain would be 1 or 2. Instead, a modulation format called complementary code keying (CCK) is used. CCK is based on the complementary codes, which have good performance with regard to mutual interference. To transmit 11 Mbps, a group of eight chips (each transmitted with one of the QPSK phases) jointly codes eight bits. Two of them are coded in the average phase rotation of the eight chips, and the other six in the selection of one out of 64 complementary codes. To transmit 5.5 Mbps, a group of eight chips jointly codes four bits. Two of them define the average phase rotation, and the other two are employed in the selection of one out of four complementary codes. These four form a subset among the 64 used for 11 Mbps. A CCK transmitter is very similar to an IEEE 802.11 DS transmitter, as the only changes are in the way the chip phases are chosen. This is typically done in the digital signal processing (DSP) part of the hardware. Therefore, a software actualization may be enough. At the receiver side, things are different, as the longer bit period of the DS receiver allows the implementation of 1- and 2-Mbps receivers with the RAKE technique. The multipath delay spread is much lower than the symbol rate. In the 5.5-Mbps and 11-Mbps version, some kind of equalization must be used, in combination with the RAKE receiver [5]. The IEEE 802.11 extension for the 5-GHz band is very similar, at the physical level, to the aforementioned HIPERLAN/2 standard. Actually, both standardization bodies (ETSI and IEEE) and the Japanese Multimedia Mobile Access Communications (MMAC) group have worked in close cooperation. IEEE 802.11 claims primacy in selecting OFDM for packetbased networks. Only minor technical details change from one to another standard. Regarding the parameters shown in Table 6.1, only the 9/16

190


coding rate used in HIPERLAN for 16-QAM is not used in the IEEE 802.11. Instead, the coding rate is 1/2. The corresponding 27-Mbps bit rate is lowered to 24 Mbps. At the MAC level there are more important differences. The MAC in the IEEE 802.11 is an extension of the CSMA/CD protocol employed in wired LANs, while HIPERLAN/2 uses a centralized MAC, based on APs and CCs [5].

6.4

Forthcoming IR standards

We can classify the new IR standards for WLANs into three main categories, the first of which includes the IrDA developments, such as IrDA OBEX, IrDA Lite, or IrDA VFIR. We also group into this category the Bluetooth-IrDA relationship. The second group includes technologies for interconnecting preexisting wired networks such as the EthIR networks for Ethernet networks or wireless access to ATM networks. The third set of developments includes new technologies for diffuse links such as the use of spread spectrum techniques. 6.4.1

IrDA new techniques

There are several new IrDA standards appearing almost every week. This is due to the fact that IrDA is an open commercial association, very flexible in adapting its products to available devices, components, or necessities. The following subsections discuss the more promising of these new standards and explore ways of increasing the range of the present IrDA transceivers. IrDA very fast IR (VFIR) [8] is a new high-speed specification, approved on March 31, 1999, that allows transmission rates of up to 16 Mbps. This represents a four-fold increase in speed from the previous maximum data rate of 4 Mbps of the IrDA FIR version. This extension provides end users with faster throughput and is backward-compatible with equipment using the current data rate. The immediate applications for these links are the interconnection of digital cameras with PCs, making it possible for users to download the entire contents of the camera in less than 20 seconds. The new standard was based on a joint proposal of the HewlettPackard Company, IBM, and the Sharp Corporation. The maximum link distance (the same as in previous versions of IrDA) is about 3m, with a

IrDA VFIR


191

±15 degree field of view. It incorporates minor changes in the IrDA-IrLAP protocol for supporting the new data rate bit. The optical components that support the VFIR specification have been demonstrated using commercially available LEDs, with a price for the 16-Mbps full implementation being about $5.00. IrDA object exchange IR object exchange (IrOBEX) is an industry standard of the IrDA that defines how objects can be shared between different IrDA devices. The IrOBEX SDK is the extended systems/counterpoints implementation of this standard and provides an application layer protocol that enables systems of all sizes and types to exchange a wide variety of data and commands in a resource-sensitive, standardized fashion. It addresses the most common application of IR device-moving arbitrary data objects to wherever the IR device is pointing. IrOBEX accomplishes this feat by packaging an object within an IrDA communications transaction, thereby enabling the transmitted object to be recognized and handled intelligently on the receiving side. IrOBEX provides the ability to put and get data objects simply and flexibly, thereby enabling rapid application development, and interaction with a broad class of devices including PCs, PDAs, data collectors, cellular phones, handheld scanners, and cameras. The application developer does not have to worry about the low-level IrDA functions of link discovery, set-up, and maintenance, but rather can focus on higher level application development. Another IrOBEX time-saving feature enables the developer to include data object information such as object description headers, along with the data object itself. The multitransport OBEX protocol is designed to allow one or more adaptation layers to provide access to various network transports. Multitransport OBEX provides a clean, well-defined interface between the OBEX protocol component and the OBEX transport adapter module. This allows easy expansion of the types of transports supported beyond those provided for IrDA IR and Bluetooth RF. Figure 6.2 shows the structure of multitransport OBEX. IrDA Lite The IrDA Lite document version 1.0 is a set of strategies and principles that complements the IrLAP and IrLMP specifications. It describes how to create a minimal IrDA implementation (minimum RAM and ROM) that is compatible with existing IrDA implementations. The basic strategies for IrDA Lite are described as follows:

192


OBEX protocol

OBEX transport interface Bluetooth transport

IrDA transport Other transports

Bluetooth stack

IrDA stack Other stacks

Figure 6.2

Structure of the OBEX protocol.

ω

To support only the required minimal link parameters, such as 9,600 bps, a data size of 64 bytes, a window size that minimizes IrLAP negotiation, and a 500-ms turnaround time;

ω

To ignore nonsupported frames in NDM and to ignore frames that do not have the correct address;

ω

To use simplified algorithms for discovery.

On the other hand, IrDA Lite does not support optional features, such as sniffing and role exchange, nor does it support UI frames (both connectionless and connection-oriented). If all the IrDA Lite strategies are applied, the result will be an implementation with the smallest possible ROM (code) and RAM (data) size. This is referred to as minimal. However, an implementation is not required to use all the IrDA Lite strategies, and any number of them can be applied. Thus, IrDA Lite implementations may range from minimal IrDA Lite through full-featured IrDA (and all points in between).


193

The weakness of the IrDA Lite implementations is that it cannot easily handle multiple independent applications running at the same time. When applications on two devices communicate, typically one of the applications (server) passively waits for the other application (client) to establish the connection. In a full-featured IrDA this situation will be handled by the IrLMP layer, but because the IrDA Lite IrLMP layer is much simpler as it does not need to manage the IrLAP layer, this is a source of potential difficulty. Potential uses of IrDA Lite include watches, printers, cameras, modems and terminal adapters, cell phones, instruments, and industrial equipment. Not all devices can live with minimal IrDA Lite capabilities but adding support for values beyond the minimum IrLAP parameters, such as higher speeds and larger data sizes can be done with reasonably small increases in RAM and ROM size. We refer to such an implementation as extended. On the other hand, IrDA Lite cannot easily support multiple applications running at the same time. However, the number of scenarios in which multiple applications need to run at the same time is small. Therefore, IrDA Lite (minimal or extended) will work for most implementations. Increasing the IrDA range According to the IrDA specification, the range is up to 1m, but in some cases it may be desired to increase link distance beyond the distance guaranteed by IrDA. The two ways of doing this are to increase transmitted light intensity or to increase receiver sensitivity. To extend the link distance, both sensitivity and intensity must be increased for both ends of the IR link. If it is desired to communicate with a standard IrDA device that may have minimum transmitter intensity, the receiver intensity must be increased. The standard IrDA device may also have minimum receiver sensitivity, so transmitter intensity must also be increased [9]. Some technical solutions have been proposed, such as the long-distance IrDA dongle, designed for the IrDA 1.0 standard [10]. As it is well known, the main problem is that the intensity drops with the square of the distance. Going from 1 to 5m requires 25 times the initial power (and battery drain on a portable device), or 25 times more sensitivity on the receiver, but maintaining the dynamic range, as it still has to be able to work at 3 inches. To do so, we have to work with a very small IrDA pulse in a severe noise environment (with fluorescent lights, screen savers, moving shadows, etc.). Another possibility relies on using laser diodes, but they are more expensive, and they are also dangerous if they

194


have more than 1 mW.1 A better solution would be to use lenses to focus the beam. [11, 12]. At first glance, it may appear that IrDA and Bluetooth [13] technologies compete with each other in the marketplace. Industry analysts have openly wondered whether both technologies can survive, given that both provide short-range wireless connectivity. Today, nevertheless, more and more people think that IrDA and Bluetooth will not only live happily side by side, but that they will also need one another. IrDA is still unbeatable in price/performance, and the new standards like AIR and VFIR will improve its possibilities even more. Bluetooth is explained in Section 6.5. It is a proposed RF specification for short-range, point-to-multipoint voice and data transfer. It has a nominal link range from 10 cm to 10m, but it can be extended to 100m by increasing the transmit power. It is based on a low-cost, short-range radio link, working and facilitating ad hoc connections for stationary and mobile communication environments. The following paragraphs compare the performance of both systems. Considering data exchange, it can be as simple as pushing a business card from a mobile phone to a PDA or as sophisticated as synchronizing personal information between a PDA and a PC. Bluetooth and IrDA specify both these applications as well as other data exchange applications. Both use the same upper layer protocol (OBEX) to implement these applications. IrDA and Bluetooth utilize the same data exchange applications when appropriate. By using the same upper-level protocol, it is possible for a single application to run on both Bluetooth and IrDA. Fortunately, the very scenarios where IrDA falls short are the ones in which Bluetooth excels and vice versa. A common data exchange scenario is one in which the exchange will take place in a room containing a number of other devices. An example is electronic business card exchange. Two people meet to exchange business cards, face-to-face, in a large conference room. Many other people carrying wireless devices are also present in the room, possibly attempting to do the same thing. This is the situation where short-range IR communications excels. Close proximity to the other person is natural in a business card exchange situation, as is IrDA and Bluetooth

1. IR technology is mostly unregulated, but due to the concern for eye safety with laser diodes and LEDs, the IEC-825-1 eye safety standard was recently amended to include all LED-based transceivers. The current IrDA specifications are well within IEC guidelines, but care must be taken as new IrDA specifications with greater range and bandwidth are created. New specifications will need to be verified against the IEC standards.


195

pointing one device at another. The limited range and angle of IrDA allows others to be carrying out a similar activity without interference. The short-range and narrow angle of IrDA provides a simple form of security and a natural ease of use. This same situation is a weakness for Bluetooth. Because of its omnidirectional characteristic, it has problems on finding the intended recipient. It does not allow the user to simply point at the intended recipient. A Bluetooth device must perform a timeconsuming search operation that will find many of the other devices in the room. Close proximity to the intended recipient will not help. The user will be forced to choose from a list of discovered devices. Choosing the proper device will probably require special information from the other person (e.g., a 48-bit device address or friendly name). Also Bluetooth has multipoint capabilities and therefore uses security mechanisms to prevent unauthorized access. The two users attempting to carry out a business card exchange would also need to execute security measures. In other data exchange situations Bluetooth is the obvious choice. It is capable of penetrating solid objects and allowing maximum mobility within a small cell. This is impossible or very difficult with IrDA. For example, with Bluetooth a person could synchronize his or her phone with a PC without taking the phone out of his or her pocket, but this is not possible with IrDA. Considering LAN access, both Bluetooth and IrDA have the ability to connect a device to a wired network wirelessly. Bluetooth has a higher level of flexibility when placing a LAN access because there are no LOS requirements. It also supports multipoint capabilities, but Bluetooths aggregate bandwidth is limited to 1 Mbps, while IrDA supports 4 Mbps, with 16 Mbps under development. With regard to voice applications, Bluetooth was designed for synchronous voice channels and has the ability to reserve bandwidth for carrying digital voice data. It supports up to three simultaneous, full-duplex voice conversations within a piconet. In the IrDA standard, the component for mobile communications is IrMC. It permits the transmission of full-duplex voice data over an IrDA linkbut consumes the full bandwidth of a 115.2 Kbps. These characteristics of Bluetooth and IrDA complement one another and demonstrate that they can (and even must) coexist. 6.4.2

Interconnection for wireless networks

Nowadays, one of the main areas of development for IR data communications is to provide a universal network access solution for portable devices, such as notebooks, handheld PCs and palm-size PCs, PDAs, and

196


the emerging Internet appliances such as smart phones and electronic books. For the notebook PC user, Ethernet is typically the preferred solution in the office due to its ability to provide high data rates, but it is rarely available when traveling. When out of the office, the business traveler with a notebook typically resorts to a dial-up modem, with at most a 56-Kbps throughput, to retrieve her or his e-mail or browse the Web. This has the added inconvenience of often having to reconfigure our programs between LAN networking (office configuration) and dial-up networking (mobile configuration). Handheld PCs, which are typically used as a companion to a desktop in the office, require a low-power PCMCIA LAN card and associated cables to connect to the corporate network. The power required by the LAN card shortens the PCs battery life; the associated cables restrict its mobility; and the cost is high. When out of the office, the business traveler with a handheld PC faces the same problems as the notebook PC user. Most palm-size PCs and PDAs do not even provide support for a LAN card and require an external modem for communication. Palm-size PCs, PDAs, and the emerging Internet appliances have no way of connecting to the Internet at high data rates. The concept of providing high-speed wireless access is based on the concept of using point-to-point protocol (PPP) over IR. It allows easy and completely wireless network access with no hardware or software to install and no reconfiguration of the portable device. Using PPP over IR, connection speeds of up to 4 Mbps are possible with the current implementation of the IrDA standard, and the recent ratification of the VFIR IrDA standard will increase this to 16 Mbps. These high-speed IR interfaces will be used in the future to bridge PCs to xDSL and cable modem for high-speed Internet access. Some other techniques, in some cases IrDA-compliant, are described briefly in the following subsections. Access to Ethernet links There are two main possibilities for providing IR access to Ethernet links. One is designing an IR interface for regular Ethernet boards. They can then connect to one another by substituting the coaxial wire or twisted pair for a LOS IR link. This possibility has become practical recently because of the availability of low-cost, high speed IREDs and photodiodes. The other one is transforming an Ethernet connection into a wireless network AP enabling LAN, WAN, and Internet access. This alternative is taken by the EthIR LAN. Both alternatives are independent of the hardware and software platform, are fully supported by the native OS, and eliminate the need for cradles, host PCs, and specific PCMCIA cards and cables.


197

EthIR, which is supported by several OSs (Windows, Linux, Apple Macintosh), provides up to 4-Mbps throughput, and is IEEE and IrDA standards-compliant. It is designed for 10Base-T up-link and auto assigns IP, DNS, and SNMP network management. The advanced version of the EthIR LAN supports multiport capabilities and is compatible with the 10/100Base-T up-link and autodetection of IR or an Ethernet connection. 6.4.3

New techniques for diffuse links: Spread spectrum

The use of diffuse wireless IR communications links for short-range indoor communications has received a lot of interest over the past few years [14]. This configuration does not need alignment. This is a very desirable advantage from the point of view of a nonexpert user. In this configuration the received power relies on the scattering on the reflecting surfaces (e.g., ceiling, walls, and furniture), making the intersymbol interference induced by multipath propagation and the high attenuation dominant issues in the maximum available baud rate. Spread spectrum techniques allow the use of low emitted power communications (under the noise level) and avoid undesired multipath propagation effects. Their performances have been tested in data applications in the IR field, but the final application of these techniques will probably be in the area of consumer electronics. Nevertheless, considerable developments still have to be made. Today, only some prototypes of direct-sequence and FH applications have been reported, mainly for the low-speed domestic environment [15]. As is well-known, spread spectrum techniques offer the possibility of avoiding the intersymbol interference induced by multipath propagation (such as in the IR indoor channel), but pose, on the other hand, higher complexities in the symbol-time recovery stage of the receiver. DSSS is based on multiplying a narrowband data signal with a baud rate R, by a broadband spreading sequence with baud rate (often known as the chip rate) Rss. As the spreading sequence is coded following a pseudonoise algorithm, only the receiver synchronized to that code would obtain the original narrowband signal. For another receiver using a nonsynchronized replica of the code, or a different sequence, the signal delivered to the decision stage will be almost white or colored noise. As the power of the original data signal is also spread throughout the coded-signal bandwidth, the power spectral density (PSD) of the coded signal will be much less than the PSD of the data signal (and even below the environmental PSD of noise). Figure 6.3 shows the utilization of the available IR spectrum for DSSS, compared with other techniques. This ensures that the

198

Wireless LAN Standards and Applications Relative Optical Emitted Power

Current Applications Remote control Serial Port IrDA

High-Speed IrDA IEEE 802.11 (Diffuse) IEEE 802.3 (LOS Prototype) DSSS (Diffuse) 1

Figure 6.3

10

25

100 Frequency (MHz)

Current IR spectrum utilization.

interference levels induced by a DSSS emitter are lower than for a nonspread system. The ratio between the baud rate and the chip rate, which is known as the process gain, defines how much the PSD of the signal is reduced (and, in the same way, how much the data bandwidth is spread). The larger the process gain, the more robust the communication against interferences or multipath propagation will become. ISI is reduced because the receiver only tracks one replica of the received signal (the signal with maximum correlation). As multipropagated replicas of the emitted signal will have delayed replicas of the code, their correlation values will therefore be lower and will be rejected (Figure 6.4). The effect of external narrowband interference is reduced because the incoming signal is also multiplied by the spread code, so the process gain divides the interference level delivered to the decision stage. As the DSSS receiver is only tuned to one code, we can establish several simultaneous communications using the same bandwidth without collision by using different codes for each link. The use of this kind of multiple access strategy (known as CDMA) is well-known in other fields of application such as cellular telephony, and offers interesting possibilities in the indoor IR channel. It avoids the use of complex protocols such as CSMA/CA or CSMA/CD, that are not well-suited to this channel as


199

lp

ls

Tx

Rx Figure 6.4 ISI induced by multipath propagation is reduced by spread spectrum techniques as it only tracks one replica of the received signal. Other replicas are delayed and its correlation with the code generated in the receiver is much lower.

DS (BPSK) - OOK (NRZ) Optical emitter

Optical receiver

R

DLL

Data Source Pseudo noise generation Clock

LPF

A

LPF

A

LPF

A

Envel. Detec. LPF A Envel. Detec. LPF A Envel. Detec. LPF A Comparison Threshold

Emitter Pseudo noise generation

VCO Sample & Conformation Receiver

Figure 6.5

+∑ -

Block diagram of the WIR-DSSS system.

ML Estimator

Data

200


they need to test the presence or absence of a carrier (and it is not easy in a carrierless system such as IR). Figure 6.5 shows the block diagram of a DSSS prototype. The main application of these systems is in the field of the communications between heterogeneous appliances in the in-house environment, such as microwave ovens, washing machines, refrigerators, and audio equipment. Currently, there is no unified system for interconnecting these systems and devices. Increasing the baud rate supported by the WIR-DSSS system will even allow the transmission of digital video and several audio channels sharing the same channel. The next section presents other RF standards.

6.5 Other RF standards: DECT, Bluetooth, WATM, HomeRF, etc. 6.5.1

Introduction

All of the systems mentioned in the title of this section have in common that they have not been (or are not being) designed specifically for WLANs. However, WLANs can be implemented with any of these technologies, and in some scenarios it may be advantageous to use one of these instead of the WLAN standards. Also, as all of them seek the economies of scale that can be reached when addressing a large number of applications, in every case cost can be an argument in favor of these technologies. Certainly, their degrees of development are quite dissimilar. DECT is the only technology that has been on the market for some time,but mostly with other applications. Bluetooth and HomeRF SWAP specifications are well-advanced. WATM is a rather general concept that includes, for example, the aforementioned HIPERLAN Type 2 and the 5-GHz extension of IEEE 802.11, but also the early developments toward the fourth generation of mobile communications. None of these systems will be described in much detail, as they are outside the scope of this book. (Strictly speaking, none of them is a wireless LAN standard.) The interested reader should consult the specialized literature or relevant Internet sites on these technologies. An extensive list of references has been included at the end of the chapter. Care has been taken in selecting the most recent and broad-scope references to lead the reader to more detailed technical papers.


201

The main reason for including these technologies in this chapter is to show the great international interest shown in providing wireless solutions for the implementation of connectivity between persons and/or devices. The etc. that is in the title includes not only those systems that are not considered because of the authors lack of awareness or the need to limit the scope of the book, but also those that, undoubtedly, will boom in the following years. 6.5.2

Digital enhanced cordless telecommunications (DECT)

DECT appeared in the early 1990s as an ETSI standard for cordless telecommunications, mainly telephony and ISDN access. Currently, the scope of DECT has been extended through the definition of DECT profiles that enable its use in a large variety of scenarios: ω

Cordless telephony (domestic or business WPABX);

ω

Cordless network access for a generic voice or data network;

ω

Wireless local loop;

ω

Cordless terminal mobility;

ω

Data services, circuit or packet switched links;

ω

Multimedia access;

ω

IMT-2000 services for fixed and low-mobility users;

ω

Narrowband WLAN.

DECT appears in this book because of the last application. Data rates can be as high as 552 Kbps in the original DECT (two-level GFSK modulation), but can be in excess of 2 Mbps with the recently approved fourand eight-level modulations. Information on the DECT standard can be obtained from the ETSI Web site [15] and from [16, 17]. A good document to start with is [18], which provides a high-level general view of the standard and a good guide to the rest of the DECT technical documents. Table 6.2 summarizes the main features of DECT. Some of the primary characteristics of the DECT standard are described as follows.

202


Table 6.2 The Main Characteristics of DECT Parameter

Values

Band

1,8801,900-MHz (Europe) 1,9001,930 MHz (other parts of the world) Other bands according to national regulations

Carriers Separation

1.728 MHz

Access

TDMA/Time Division Duplex (TDD) in 10-ms frames

Modulation (original)

GFSK with BT = 0.5

Modulation (extension)

(π/2-DBPSK, (π/4-DQPSK, and (π/8-D8PSK

Symbol rate

1.152 Ms/s

Bit rate

1.152, 2.304, and 3.456 Mbps

Voice channels

12 (duplex) at 32 Kbps per carrier

Max. Transmit Power

250 mW

Receiver Sensitivity

−86 dBm

Channel assignment

Dynamic channel selection

Other features

Authentication, encryption, seamless handover

ω

DECT terminals are classified as fixed terminals (FTs) and portable terminals (PTs). An FT defines a cell, and PTs within a cell synchronize their transmission according to the time base defined by the FT. Physical channels are defined in the frequency/time scales, according to an FDMA/TDMA/TDD scheme. Direct communication between two PTs is allowed, but one of them acts as an FT in this case. Wireless relay stations (WRSs) allow the extension of an FT radio range. The WRS acts as a PT, when seen from the FT, and as an FT when seen from the PTs.

ω

A carrier frequency and a slot number define a physical channel. In Europe there are 10 carriers in the 1,8801,900-MHz frequency range. In other parts of the world, frequencies between 1,900 and 1,930 MHz may be used. The use of other microwave bands for fixed wireless access, possibly with transmit power larger than 250 mW, is contemplated in the standard.

ω

Frame duration is 10 ms, and a frame is structured in 24 full slots. If traffic is symmetric, 12 are employed for each link, the first ones


203

(011) for the downlink (FT to PTs) and the last ones (1223) for the uplink (PTs to FT). There are also half slots (with half duration) and double slots. A full slot typical application is telephony at 32 Kbps (unprotected) or data communications at 24 Kbps (protected). However, several slots can be used in a communication to increase the bit rate. With the original GMSK modulation, a maximum of 23 slots can be assigned for one-way data communications, with a net data rate of 552 Kbps and very low BER(,10−9). With eight-level modulation, the maximum data rate (unidirectional) is in excess of 2 Mbps. ω

Control information and user data are multiplexed in the slots, however long its duration.

ω

The system includes dynamic channel selection. The portable terminals select the carrier and time slot that are best suited for the communication. This can be done at the start of the communication but also throughout its course, particularly if the channel is degraded because of interference or propagation impairments. Seamless handover is applied.

ω

With regard to the LAN application, the data packet radio service (DPRS) defines features and services common to all packet data applications, with and without mobility. Frame relay services are supported for Ethernet, token ring, IP, and PPP applications. Wireless terminals can make up an ad hoc network or be connected to a wired network.

ω

WLANs using DECT enjoy all the advantages of this standard, with regard to the efficient use of the spectrum, cellular configuration with intracell and intercell handover, authentication and data encryption, spectrum availability, and coexistence with other networks in the same band. More than 45 million DECT terminals are installed worldwide, making it the most successful cordless standard. Consequently, DECT terminals are comparatively very cheap. The main disadvantage is the relatively low data rate (552 Kbps) provided by the original DECT. This will improve with the new equipment that includes the higher-level modulation formats.

204

6.5.3


Bluetooth

Bluetooth is an initiative of five major manufacturers from the computer and cellular communications fields: Ericsson, Nokia, IBM, Toshiba, and Intel. Version 1.0 A was published in 1999 and includes a core [19] and a set of profiles [20] that are related to applications. All Bluetooth devices implement the core. It is necessary to implement the profiles needed for each application. Version 1.0 B of the specifications has been recently released, and can be obtained from [21]. A very good high-level introduction to Bluetooth is found in [22]. Bluetooth enables the easy and cheap wireless interconnection of electronic devices of any kind. Communication is limited to the vicinity of the devices and need not be structured in networks with ad hoc configurations. On the contrary, Bluetooth networks, named piconets, are made up and canceled spontaneously according to the services needed. Some of the key applications include the connection of a laptop computer to a mobile phone (for example, to send data through a cellular network), the connection of a mouse to the PC, or the implementation of wireless headsets for mobile phones. The range of applications is not limited, nor is the profiles list closed. New profiles can be added according to

Table 6.3 The Main Characteristics of Bluetooth Parameter Band

Values 2.45 ISM band Different channels according to the country (see Chapter 3)

Carriers Separation

1 MHz

Access

FH; TDD

Modulation

GFSK with BT = 0.3

Bit rate

1 Mbps

Voice channels

64 Kbps

Maximum transmit power

1 mW nominal 100 mW with closed loop power control

Receiver sensitivity

< −70 dBm (tentative)

Other features

Authentication, encryption, power-saving functions, interpiconet communications


205

newly identified needs. Table 6.3 lists the main characteristics of Bluetooth. The Bluetooth specification provides low-cost connectivity by using a PHY specification with relaxed technical features compared to other systems. Also, as economies of scale are very important to reduce manufacturing costs, the first version of Bluetooth works in the ISM 2.4-GHz band (as does IEEE 802.11), which is available worldwide, although there are some restrictions in a number of countries. FH is employed to allow the coexistence of Bluetooth piconets with other transmissions of the same kind, or of other systems. In the scenarios for which Bluetooth is promoted, a large number of piconets can coexist in the same location without coordination. For example, inside the same room of an office, several ad hoc piconets may link the mice, keyboards, and printers with the computers. At the same time, some people may be talking with a wireless headset that is connected to their mobile phones, while others are transferring data from the PC to the mobile phone to send it through the GSM network. One device may participate simultaneously in more than one piconet (interpiconet communications). Figure 6.6 presents the concept of scatternet or combination of piconets that is proper to Bluetooth. The gross bit rate is 1 Mbps, with GFSK modulation. The hop sequences are selected on the basis of user identity and are not orthogonal but have a low probability of persistent interference. There is a

Figure 6.6

The combination of piconets in Bluetooth.

206


TDMA/TDD structure linked to the hopping pattern. The slot duration is 625 µs. The packet length is equal to one, three, or five slots. The carrier frequency is fixed within a packet but changes between subsequent packets. The hop sequence in a piconet is provided by the identity of one device, which acts as master. It also provides the phase of the sequence. The remaining devices are slaves and coordinate their transmissions to the hop sequence and phase of the master. Any device can act as master or slave. Usually, the one that requests the service acts as master, except when the net is already established. High-quality audio at 64 Kbps can be provided through the use of a synchronous connection-oriented (SCO) link. Duplex slots (two continuous slots, one for each direction) are reserved at regular intervals. The remaining slots can be used by asynchronous connectionless (ACL) links, scheduled by the master. With regard to the applications in WLANs, Bluetooth core includes the OBEX protocol, for interoperability with IrDa. There is also a LAN access profile. It defines how Bluetooth devices can access LAN services through a LAN AP, with PPP. Also, two Bluetooth devices can communicate using PPP as if they were part of a LAN. However, Bluetooth does not aim to establish a complete LAN. Based on Bluetooth, the IEEE P802.15 working group is preparing a standard for wireless personal area networks [23]. It is addressed to applications that need to communicate via devices that are around a person. These are the same applications that are the focus of Bluetooth. 6.5.4

Wireless ATM

ATM is one of the leading technologies in fixed high-capacity networks. In most situations, ATM is implemented in optical fiber links, cables, or fixed microwave point-to-point links. The concept of WATM relates to the extension of ATM services to other scenarios, through the use of wireless transmission and featuring mobility. It includes the wireless mobile ATM, which is the basis for providing services in the order of tens of megabits per second to mobile users, satellite ATM (where the large delays are significant), and WLANs. In the last few years, there has been much international activity in these fields. A large number of WATM projects have been launched. The interested reader can refer to [24], which, though dedicated to the ATMmobil Project, is a good introduction to the applications of WATM, with an extensive list of references. For a reference to the current


207

research topics in this area, there are several papers in the IEEE Communications magazine, vol. 37, no. 11, Nov. 1999. With regard to WLANs, both HIPERLAN type 2 and the IEEE 802.11 extensions to HBRs can provide ATM services [5], as has already been mentioned. Also the MMAC project has the objective of delivering ultrahigh-speed data rates to, for example, WLANs, through the use of WATM. 6.5.5

HomeRF

Around 100 manufacturers from the computers, communications, and microelectronics fields make up the HomeRF working group. The specification has been prepared to provide wireless voice and data networking in the home. The expected radio range is on the order of 50m indoor (1020m for low-power devices). The specified radio access is called SWAP. Some examples of applications of HomeRF are described as follows: ω

Set up a wireless home network to share voice and data between PCs, peripherals, PC-enhanced cordless phones, and new devices such as portable, remote display pads;

ω

Access the Internet from anywhere in and around the home from portable display devices;

ω

Share an ISP connection between PCs and other new devices;

ω

Share files/modems/printers in multiPC homes;

ω

Intelligently forward incoming telephone calls to multiple cordless handsets, FAX machines, and voice mailboxes;

ω

Review incoming voice, FAX, and e-mail messages from a small PCenhanced cordless telephone handset;

ω

Activate other home electronic systems by simply speaking a command into a PC-enhanced cordless handset;

ω

Play multiplayer games and/or toys based on PC or Internet resources.

Table 6.4 lists the main technical features of SWAP.

208


Table 6.4 The Main Characteristics of SWAP Parameter

Values

Band

2.45 ISM band; different channels according to the country (see Chapter 3)

Carriers separation

1 MHz

Access

FH; TDMA/TDD

Modulation

2-FSK, 4-FSK

Symbol rate

0.8 Ms/s

Bit rate

0.8 Mbps, 1.6 Mbps

Voice channels

Four channels at 32 Kbps

Maximum transmit power

100250 mW nominal 12.5 mW low power devices

Receiver sensitivity

−80 dBm (2-FSK)

Other features

USB connection to the PC; connection to the PSTN

The PHY specification is based on the IEEE 802.11 FH mode, with data rates of 0.8 and 1.6 Mbps and a hop time of 300 µs. The frame duration, equal to the hop time, is structured in two parts, called subframes: ω

A TDMA/TDD subframe intended for isochronous communications, mainly voice communications. A maximum of 4 simultaneous voice communications can be carried, with slots reserved for retransmissions. The voice link is based on DECT and uses 32 Kbps ADPCM high-quality voice coding.

ω

A CSMA/CA subframe based on IEEE 802.11 but with some of the more costly features removed. This is used for peer-to-peer asynchronous data communications, making up an effective WLAN.

A HomeRF connection point governs the net, and is connected to a PC (typically with Internet access) and to the PSTN. HomeRF asynchronous devices can connect to each other without the intervention of the control point. The control point is necessary for voice communications, which can be streamed to the PSTN or be established between two users within the net. Figure 6.7 shows a representation of a HomeRF network.


209

Asynchronous connections

HomeRF control point

PSTN

USB Isochronous links Figure 6.7

Internet access

An example of HomeRF network.

Enhanced cordless telecommunications is provided by the presence of the PC, which can add features, for example, to route the incoming calls to a specific handset (based on the caller ID) or to store and generate voice messages. The HomeRF-based WLAN cannot compete in the industrial and business fields with the 802.11 and HIPERLAN standards, because of its low range and reduced features. However, it is a good candidate for domestic LANs that only integrate a reduced number of devices (e.g., fixed PC, laptop, printer, and digital camera) in a limited physical range. The complete specifications of HomeRF can be obtained from [25]. A good description of the system has been recently published [26].

6.6

Conclusions

New wireless applications are defined continuously, and a large number of RF and IR standards and specifications are being proposed to satisfy users needs. Not all of them will survive in the medium term, and it is difficult to say which are in the best position. However, it is widely accepted that wireless systems for data communications, WLANs among them, will show an impressive growth in coming years. If this book can serve the reader to better understand and take advantage of the broad range of opportunities brought about by these systems, the authors will be satisfied.

210


References [1]

Bhagavath, V. K., Emerging High-Speed xDSL Access Services: Architectures, Issues, Insights, and Implications, IEEE Comm., Vol. 37, No. 11, Nov. 1999, pp. 106114.

[2] http://www.hiperlan.com. [3] ETSI TR 101 683, Broadband Radio Access Networks (BRAN); HIPERLAN Type 2; System Overview, Feb. 2000. [4] http://www.hiperlan2.com. [5] Van Nee, R., et al., New High-Rate Wireless LAN Standards, IEEE Comm., Vol. 37, No. 12, Dec. 1999, pp. 8288. [6] Vornefeld, U., C. Walke, and B. Walke, SDMA Techniques for Wireless ATM, IEEE Comm., Vol. 37, No. 11, Nov. 1999, pp. 5257. [7] http://www.etsi.org/BRAN. [8] The IR Data Association Approves High Speed Extension to IrDA Increasing Current Transmission Rates Four-Fold, http://www.irda.org. [9] IrDA Data Link Design Guide, Hewlett-Packard, http://www.hp.com/go/ir. [10] ELEKTOR 5/97, http://www.elektor.de. [11] Pérez-Jiménez, R., V. M. Meliàn, and M. J. Betancor, Analysis of Multipath Impulse Response of Diffuse and Quasi-Diffuse Optical Links for IR-WLAN, Proceedings IEEE INFOCOM 95, Boston, MA, Apr. 1995. [12] Gabiola, F., et al., Irradiance Analysis for Indoor Point-to-Point and Q-Diffuse IR Channels, Microwave and Optical Tech. Letters, Vol. 6, No. 9, July 1993. [13] http://www.etsi.org/DECT. [14] Kahn, J., and J. Barry, Wireless Infrared Communications, Proceedings of the IEEE, Vol. 85, No. 2, Feb. 1997. [15] Vento, R., et al., Experimental Characterization of a Direct Sequence Spread Spectrum System for In-House Wireless Infrared Communications, IEEE Transactions on Consumer Electronics, Nov. 1999. [16] http://www.dect.ch. [17] http://www.dectweb.com. [18] ETSI TR 101 178, Digital Enhanced Cordless Telecommunications (DECT); A High Level Guide to the DECT Standardization, Mar. 2000. [19] Specification of the Bluetooth System, Core, V. 1.0 A, July 1999. [20] Specification of the Bluetooth System, Profiles, V. 1.0 A, July 1999.


211

[21] http://www.bluetooth.com. [22] Haartsen, J. C., The Bluetooth Radio System, IEEE Personal Communications, Vol. 7, No. 1, Feb. 2000, pp. 2836. [23] Siep, T. M., et al., Paving the Way for Personal Area Network Standards: An Overview of the IEEE P802.15 Working Group for Wireless Personal Area Networks, IEEE Personal Communications, Vol. 7, No. 1, Feb. 2000, pp. 3743. [24] Keller, R., et al., Wireless ATM for Broadband Multimedia Wireless Access: The ATMmobil Project, IEEE Personal Communications, Vol. 6, No. 5, Oct. 1999, pp. 6680. [25] http://www.homerf.org. [26] Negus, K. J., A. P. Stephens, and J. Lansford, HomeRF: Wireless Networking for the Connected Home, IEEE Personal Communications, Vol. 7, No. 1, Feb. 2000, pp. 2027.

.

Glossary ACK

acknowledgment

ACL

asynchronous connectionless

ADA

alias destination address address field

ADDR

asymmetric digital subscriber loop

ADSL AIB

alias information base

AID

acknowledgment identifier field

AP

access point

ARQ

automatic repeat address

ASA

alias source address

ATM

asynchronous transfer mode abstract test suite

ATS

additive white Gaussian noise

AWGN BCH

Bose-Chaudhuri-Hocquenghem

BER

bit error ratio

BLI

block length indicator

BLIR BLIRCS

block length indicator replica BLIR checksum 213

214


Bluetooth SIG

Bluetooth special interest group

business premises network

BPN

broadband radio access network

BRAN BSA

basic service area

BSS

basic service set Broadband Wireless Internet Forum

BWIF

channel access control

CAC

central controller

CC CCA

clear channel assessment

CCK

complementary code keying code division multiple access

CDMA CF

coordination function

CFP CP

contention-free period contention period

CP-HCPDU

channel permission HCPDU

CPU

central processing unit

CRC

cyclic redundancy checking

CSMA

carrier sense multiple access

CSMA/CA

carrier sense multiple access with collision avoidance

CSMA/CD

CSMA with collision detection

CTS

clear to send

CW

contention window

DBP

groups of two bits that will be 4PPM coded (data bit pair)

DCF

distributed coordination function

DCLA DD DECT

DC level adjustment data symbols after 4PPM codification of a pair of bits digital enhanced cordless telecommunications

Glossary

215

data link layer

DLL

downlink

D-link

direct-link

Di-link DPN

domestic premises network

DPRS

data packet radio service

DR

data rate

DS

distribution system

DSL

digital subscriber loop

DSP

digital signal processing

DSSS

direct sequence spread spectrum

DTBS

distributed time-bounded services

ED

energy detect equivalent radiated isotropic power

EIRP

effective isotropic radiated peak envelope power

EIRPEP EMC

electromagnetic compatibility

EMI

electromagnetic interference

ESS

extended service set

ETSI

European Telecommunications Standards Institute

EY-NPMA access

elimination yield nonpreemptive priority multiple

FCS

frame check sequence

FEC

forward error correction

FER

frame error rate

FFH

fast frequency hopping

FH

frequency hopping

FHSS FPLMTS

frequency hopping spread spectrum future public land mobile telecommunications system

216


FSK

frequency shift keying

FSM

finite state machines fixed terminal

FT

Gaussian frequency shift keying

GFSK

Gaussian minimum shift keying

GMSK

global system for mobile communications

GSM

GP-HMPDU

group-attention pattern HMPDU

high bit rate

HBR HCPDU

HIPERLAN CAC protocol data unit

HCSAP

HIPERLAN CAC service access point

HCSDU

HIPERLAN CAC service data unit hashed destination address

HDA HDLC

high-level data link control

HDSL

high-speed digital subscriber loop

HEC HI

header error check HBR-part indicator

HIB

hello information base

HID

HIPERLAN identifier high-performance radio local area network

HIPERLAN

HIPERLAN MAC protocol data unit

HMPDU HMS-user

Ho-HMPDU HPC

HIPERLAN MAC service user hello HMPDU

handheld personal computer

IAP

information access protocol

IAS

information access service

IBCN IBSS

integrated broadband communications network independent BSS

Glossary

217

interframe space

IFS

individual-attention pattern HMPDU

IP-HMPDU IR

infrared IR communications protocol

IRCOMM

Infrared Data Association

IrDA

IrDA plug and play

IrDAPNP

infrared emitting diode

IRED IrLAP

infrared link access protocol

IrLMP

infrared link management protocol infrared transference protocol

IrTRAN-P ISAP

IrLAP services access point

ISDN

integrated services digital network

ISM ISP IT ITU IV

industrial, scientific, medical Internet service provider information technology International Telecommunications Union initialization vector

KID

key identifier

LAN

local area network

LBR

low bit rate

LC-HMPDU LI

look-up confirm HMPDU

length indicator

LLC

logical link control

LME

physical layer management entity

LM-IAS

link management information access service

LM-MUX

link management multiplexor

LM-PDU

LM-MUX protocol data units

218


LOS

line of sight

LPS

local position system

LR-HMPDU

look-up request HMPDU

LM-MUX services access point

LSAP

LSAP-SEL

LSAP selector

least significant bit

LSB

medium access control

MAC

maximum adaptive defer threshold

MADT

mobile broadband system

MBS

MSDU lifetime

ML

MMAC

multimedia mobile access communications

MPDU

MAC protocol data unit

MSAP

MAC service access point

MSDU

MAC service data unit

MSB

most significant bit

MSN

multipoint relay set sequence number microwave

MW

neighbor address

NA NAV

net allocation vector

NDM

normal disconnect mode

NIB

neighbor information base

NIC

network interface controller normalized residual lifetime

NRL

normal response mode

NRM NS

neighbor status

OA

originator address

OFDM

orthogonal frequency division multiplex

Glossary

219

open system interconnection

OSI

preamble field

PA

point coordination function

PCF PDA

personal digital assistant

PDC

personal digital cellular

PDU

protocol data unit poll/final

P/F PHY

physical layer

PIFS

point-CF IFS physical layer convergence protocol

PLCP

padding length indicator

PLI PLW

PLCP-PDU length word

PMD

physical medium dependent

PPM

pulse position modulation

PSD

power spectral density

PSF

PLCP-PDU signaling field public-switched telephone network

PSTN PT

portable terminal

QAM QoS

quaternary amplitude modulation quality of service

QPSK

quadrature phase shift keying

RACH

random access channel

RES RF RIB RL RMS

radio equipment and system radio frequency route information base residual lifetime root mean squared

220


RTI

relay type indicator

RTS

request to send

RX

receiver

SA

source address Security Algorithm Group of Experts

SAGE

segmentation and reassembly

SAR SC

sanity check synchronous connection-oriented

SCO

symmetric digital subscriber loop

SDSL SFD

start frame delimiter

SFH

slow frequency hopping

SIFS

short IFS

SIP

serial infrared interaction pulse

SIR

serial IR

SLN

signal level number source multipoint relay address

SMA

source multipoint relay information base

SMRIB STA

start flag

STO

end flag

S-UMTS

satellite UMTS

SWAP

shared wireless access protocol

TDMA

time division multiple access

TIB

topology information base

TSS&TP

test suite structure and test purposes

TTP-PDU

TinyTP protocol data units

TTPSAP

TinyTP services access point

T-UMTS

terrestrial UMTS

Glossary

TX

221

transmitter

UART

universal asynchronous receiver-transmitter

U-link

uplink

UMTS/IMT-2000 universal mobile telecommunication system/ international mobile 2000

unlicensed national information infrastructure

UNI UP

user priority very-high-speed digital subscriber loop

VDSL

very-fast IR

VFIR

wireless ATM access systems

WACS WAND

wireless adaptive network device

WATM

wireless ATM

WECA

Wireless Ethernet Compatibility Alliance

WEP

wired equivalency privacy wireless LAN

WLAN WLANA WLIF

Wireless LAN Association Wireless LAN Interoperability Forum

WLL

wireless local loop

WRS

wireless relay station

.

About the Authors Asunción Santamaría ([email protected]) was born in Madrid, Spain, in 1965. She received her telecommunications engineering degree and her Ph.D. from the Universidad Politécnica de Madrid (UPM) in 1988 and 1994, respectively. She is a professor of circuits analysis at the E.T.S.I. Telecomunicación of UPM. Professor Santamarías research interests are in the areas of optical wireless communications systems. She is a member of the IEEE. Francisco. J. López-Hernández ([email protected]) was born in Cartagena, Spain, in 1955. He received his Ph.D. in telecommunications engineering from UPM in 1983. He is a professor of optical communications at the E.T.S.I. Telecomunicación of UPM. He was in the IEEE 802.11 working group. His research of the last 20 years has been devoted to wireless optical communications. López-Hernández is a member of the IEEE LEOS, Communication, and Computer societies and of SPIE. Rafael Pérez-Jiménez ([email protected]) received his M.S. in 1991 from UPM, and his Ph.D. in 1995 from the Universidad de Las Palmas de Gran Canaria (ULPGC), both in Spain. He is now a professor at the Universidad de Las Palmas de Gran Canaria. His current research interests are in the field of wireless LANs and the design of IR systems for indoor communications. Victor Melián ([email protected]) received his telecommunications engineering degree and Ph.D. in telecommunications engineering from ULPGC in 1992 and 1997, respectively. From 1992 to 1996 he carried out 223

224


research and development of optical communication systems at ULPGC. From 1996 to 1999 he served as an assistant professor in the Department of Signals and Communications at ULPGC. He is currently a professor of optical communications at E.T.S.I. Telecomunicación of ULPGC. His research interest areas are wireless infrared communications systems and digital systems. Juan C. Miñano ([email protected]) received his telecommunications engineering degree and Ph.D. from UPM in 1982 and 1986, respectively. He is a professor at E.T.S.I. Telecomunicación at the Universidad Politécnica de Madrid. His research interests are in the areas of optics applied to optical wireless communications, photovoltaic concentration, and LED illumination systems. José A. Rabadán received his M.S. in 1995 and his Ph.D. in 2000, both from ULPGC, Spain. He is currently working on wideband CDMA systems for IR indoor channels. José M. Riera ([email protected]) received his telecommunications engineering degree and Ph.D. from UPM in 1987 and 1991, respectively. Since 1993 he has been a professor of radiocommunications at E.T.S.I. Telecomunicación of UPM. His research interests are in the areas of radiowave propagation and wireless communications systems. He is a member of the IEEE. José R. Vento-Álvarez ([email protected]) received his telecommunications Engineering degree from the Instituto Superior Politécnico José A. Echevarria (Havana City) in 1982. He received his M.S. and Ph.D. from UPM, Spain, in 1995 and 1998, respectively. He is a professor of telecommunications systems at the University of Pinar del Río (Cuba). His research interests are in the area of optical wireless communications systems.

Index Airport scenarios, 15758 Alias information base (AIB), 137, 144, 145 Application scenarios, 15181 business environment, 15859 conclusions, 17677 domestic buildings, 15960 education, 15355 health care, 15556 high-security buildings, 15657 industrial, 16061 introduction, 15152 public buildings, 1521589 public transportation, 15758 technologies and products, 16176 types of, 151 Asymmetric DSL (ADSL), 3 Asynchronous transfer mode (ATM), 3 ATMmobil Project, 206 Base-hopping sequences, 84, 85 France, 85 North America/Europe, 84 Spain, 85 Basic service area (BSA), 54 Basic service sets (BSSs), 54, 56 Battery power consumption, 4950 Bit ordering, 102 Bluetooth, 2, 47, 2036 characteristics, 204

choice of, 195 combination of piconets in, 205 defined, 194, 2034 devices, 204 IrDA and, 19495 LAN access and, 195 OBEX protocol, 206 voice applications and, 195 weakness, 195 Bluetooth SIG, 162, 166 BPSK, 79, 80 Broadband ISDN (B-ISDN), 3 Broadband radio access networks (BRANs), 5 ETSI, 168 standards, 14748 Broadband Wireless Internet Forum (BWIF), 162, 16768 adopter members, 168 defined, 162 promoter members, 16768 Business environment scenarios, 15859 advantages, 158 example, 15859 needs and, 158 See also Application scenarios Cabled networks, 184 CAC, 111, 12131 channel access, 12426

225

226

CAC (continued) channel access in free channel condition, 12627 channel access in synchronized channel condition, 12730 channel-free interval, 126 defined, 121 factors for collision avoidance, 12526 free channel, 124 generalities, 12122 HCPDUs, 12224 hidden elimination, 125 hidden node detection and operation, 13031 objectives, 12122 responsibilities, 122 synchronized channel, 124, 131 synchronized channel condition, 12730 See also HIPERLAN standard Central limit theorem, 82 CF-aware stations, 63 Channel access control. See CAC Code division multiple access (CDMA), 5 Complementary code keying (CCK), 189 Contention-free period (CFP), 63 maximum size, 63 repetition interval, 64 CSMA/CA, 57, 58, 198 with four-way handshake, 59 random backoff procedure, 62 CSMA/CD, 58, 198200 Cyclic redundancy check (CRC), 16 16-bit, 16 32-bit, 1920 bits representation, 20 calculation, 16, 20 fields, 17 Data link layer (DLL), 45 LLC, 45 MAC, 45 Data packet radio service (DPRS), 203 Delay, 48


Delay lock loop (DLL), 80, 81 Diffuse links, 197200 spread spectrum for, 197200 use of, 197 Digital enhanced cordless telecommunications (DECT), 4, 2013 characteristics, 202 control information, 203 data rates, 201 DPRS, 203 dynamic channel selection, 203 frame duration, 202 scenarios, 201 standard information, 201 terminals, 202, 203 user data, 203 Digital subscriber loop (DSL), 3 Direct sequence spread spectrum (DSSS), 197200 basis, 197 emitter, 198 prototype block diagram, 199 receivers, 198 Direct sequence systems, 7583 advantages, 77 codes in, 7879 delay lock loop, 80, 81 modulation rate, 75 orthogonal codes, 7879 performance in, 8283 power spectral density, 77 quasi-orthogonal codes, 79 reasons for using, 77 receiver, 76 receiver performance, 82 receiver structure, 81 transmission and reception in, 7992 transmitter, 76 Distributed coordination function (DCF), 5758 CSMA/CA basis, 57 defined, 56 with handshaking, 56 operation, 57 See also IEEE 802.11 standard

Index Domestic building scenarios, 15960 DS-SS physical layer, 69, 8993 bandwidth, 95 channels, 90 clear channel assessment (CCA), 91 data descrambler, 94 data scrambler, 94 FH-SS comparison, 9396 functional entities, 92 isolated network operation, 95 modulation encoding, 91 MPDU length, 92 PLCP fields, 9293 PLCP frame format, 93 processing gain, 90 protection, 95 protocols and functions, 9293 radio transmission, 8991 receiver sensitivity, 91 transmission modes, 89 transmitter power, 91 transmit-to-receive switching time, 9091 See also IEEE 802.11 standard DT-HMPDU, 139, 14041 defined, 139 residual lifetime (RL), 14041 structure, 141 type indicator, 140 See also HIPERLAN MAC PDUs (HMPDUs) Education scenarios, 15355 EIRIS Healthcare, 17475 Elimination-yield nonpreemptive priority multiple access (EYNPMA), 111 ELPAS AIRnet, 17576 Ethernet, 196 link access, 19697 See also Wireless Ethernet Compatibility Alliance (WECA) EthIR, 19697 advanced version, 197 support, 197

227

ETSI BRAN, 168 ETSI HiperLAN, 168 Fast frequency hopping (FFH), 72 FH-SS physical layer, 69, 8389 base-hopping sequences, 84, 85 channel switching, 86 convergence function, 86 DS-SS comparison, 9396 fading possibility, 95 frequency bands, 83 GFSK modulation, 85 isolated network operation, 95 layer management entity (LME), 87 minimum hop size, 84 MPDU length, 88 physical medium-dependent function, 87 PLCP frame format, 8889 PLCP sublayer, 87 PMD sublayer, 87 protocols and functions, 8689 radio transmission, 8386 receiver threshold, 86 reference model, 87 reference receiver sensitivity, 86 transmission modes, 83 transmitter center frequency accuracy, 86 transmitter power, 86 See also IEEE 802.11 standard Frequency hopping systems, 7275 coherence bandwidth, 74 fast (FFH), 72 modulation, 72 pass-fail operation, 73 receivers, 75 scenario, 75 slow (SFH), 72 transmission, 73 transmitters, 75 Future, 183209 fixed access, 185 inside home, 185 Internet access, 185 Is it wireless? question, 18386

228

Future (continued) telephony, 185 trends, 37, 183209 wireless, features, 18485 Gaussian frequency shift keying (GFSK), 85, 116 advantages, 116 use of, 85 Global system for mobile communications (GSM), 4 GP-HMPDU, 140 defined, 139 structure, 141 See also HIPERLAN MAC PDUs (HMPDUs) Handheld PCs, 6, 196 Health care scenarios, 15556 example, 15556 medications in, 156 personal communication devices in, 155 See also Application scenarios Hello information base (HIB), 13637, 144, 145 High-level data link control (HDLC), 23 High-performance radio LAN. See HIPERLANs; HIPERLAN standard High-security buildings scenarios, 15657 High-speed DSL (HDSL), 3 HIPERACCESS, 147 Hiperlan 1 Global Alliance (H1GA), 168 HIPERLAN/2, 147, 148, 18688, 200 basis, 186 in centralized mode, 186 OFDM parameters, 188 radio channels, 187 reference model, 187 standardization process, 188 Hiperlan 2 Global Alliance (H2GA), 168


HIPERLAN CAC PDUs (HCPDUs), 12224 AK-HCPDU, 122, 123 CP-HCPDU, 122, 123 DT-HCPDU, 123, 124 HID field, 124 LBR-HBR-HCPDU, 122 LBR-HCPDU, 122 types, 122 HIPERLAN MAC PDUs (HMPDUs), 111 delivered, 112 DT-HMPDU, 139, 14041 GP-HMPDU, 139, 140 HO-HMPDU, 139, 14142 IP-HMPDU, 139, 140 LC-HMPDU, 13940 LR-HMPDU, 13940 TC-HMPDU, 139, 14142 types of, 139 HIPERLANs ad hoc, 110 channel properties, 11011 defined, 110 EYNPMA, 111 features, 111 high bit rate (HBR), 111 identifier, 132 low bit rate (LBR), 111 name, 132 power-saving functions, 111 predefined, 110 routing procedures, 111 HIPERLAN standard, 2, 10948 area, 3 bit rates, 110 CAC sublayer, 111, 12131 channel access, 67 communication model, 113 defined, 10910 environment, 34 evolution, 18688 introduction, 10913 MAC protocol comparison, 6669 MAC sublayer, 111, 13140 packet transmission mechanism, 68

Index PHY, 111, 11321 prioritization phase, 67 reference model, 112 technical committee, 68 type 1, 110, 146, 186 type 2, 147, 148, 18688 HIPERLINK, 14748 HO-HMPDU, 14142, 143, 144 defined, 139 structure, 142 See also HIPERLAN MAC PDUs (HMPDUs) HomeRF, 160, 162, 16667, 2079 alliance, 167 application examples, 207 connection points, 208 consumer applications, 16667 defined, 162 example network, 209 Web site, 166, 167 working group, 167, 207 Home scenarios, 15960 benefits, 159 equipment, 159 installation, 160 See also Application scenarios IEEE 802.11 standard, 2, 45105 applications, 1045 area, 3 CCK, 189 complexity, 46 conclusions, 1045 coordination functions, 56 defined, 46 direct sequence physical layer, 8993 direct sequence systems, 7583 disadvantages, 146 distributed coordination function (DCF), 5758 DS-SS physical, 69 environment, 34 evolution, 18890 extensions, 18889 FH-SS physical layer, 69

229

frequency hopping physical layer, 8389 frequency hopping techniques, 7275 general description, 4547 goal, 46 IR-PLCP, 9798 IR-PMD, 99104 layer structure, 46 point coordination function (PCF), 56, 6366 regulatory issues, 71 RTS-CTS procedure, 5863 spread spectrum techniques, 7172 support, 47 time slots, 62 Web site, 168 IEEE 802.11 WLANs applications, 1045 basic service area (BSA), 54 basic service sets (BSS), 54 MAC for, 4769 physical layer (infrared systems), 96104 physical layer (radio systems), 6996 IEEE 802.15, 168 IEEE 802.3, 45 Industrial scenarios, 16061 benefits, 160 examples, 161 See also Application scenarios Infrared Data Association. See IrDA Infrastructure LANs, 51 Integrated broadband communications network (IBCN), 5 Integrated serviced digital network (ISDN), 3 Interconnection, 19597 Ethernet links, 19697 handheld PCs, 196 Interference, 52 Interframe space (IFS), 59 International mobile telecommunication-2000 (IMTS-2000), 5

230

Internet access, 185 IP-HMPDU, 140 defined, 139 structure, 141 See also HIPERLAN MAC PDUs (HMPDUs) IR communications protocol (IrCOMM), 1011 IrDA, 2, 942 area, 3 bit rates, 13 bit rates from 2.4 to 115.2 Kbps, 1415 bit rates of 0.576 and 1.152 Mbps, 1517 bit rates of 4 Mbps, 1721 Bluetooth and, 19495 defined, 9 environment, 34 focus, 2 frames, 14, 15, 16 goal, 9 Infrared Dongle Interface, 12 IrCOMM, 1011, 12 IrLAN, 12, 3842 IrLAP, 12, 16, 2228 IrLMP, 12, 2835 IrTRAN-P, 12 new techniques, 19095 object exchange, 191 one-end serial transmitter, 14 optical interface characteristics, 2122 physical layer, 1322 protocol stack, 10, 1113 range, increasing, 193 Serial Infrared Physical Layer Measurements, 11 signaling rate, 17 SIR interaction pulse (SIP), 15 structure, 1113 TinyTP, 12, 3537 version 1.0, 10 version 1.1, 1011 VIFR, 19091 Web site, 9, 169 IrDA companies, 16976


components manufacturers, 170 desktop computers/peripheral devices products, 172 mobile computing products, 171 service providers, 175 software providers, 17374 IrDA Lite, 12, 19193 defined, 191 implementations, 19293 potential uses, 193 support, 192 weakness, 193 IrDA membership, 17881 IrLAN, 3842 access methods, 3941 access point mode, 39 command description, 42 control channel frames format, 42 data frame format, 4142 defined, 12 features, 38 frames format and size, 4142 general description, 38 hosted mode, 40, 41 operation modes, 39 peer-to-peer mode, 40 See also IrDA IrLAP, 16 basis, 2223 configurations and operating characteristics, 2526 connectionless services, 24 connection-oriented services, 2425 data link states, 25 data stations, 25 defined, 12, 22 frame sequencing, 27 frame structure, 26 frame types, 2628 information (I) frames, 27 as mandatory level, 28 modes, 2526 poll/final (P/F) bit, 2728 service and primitives structure, 24 service primitives, 23 services access points (ISAP), 29

Index service types, 23 supervisory (S) frames, 27 unnumbered (U) frames, 27 wrapping layer, 26 See also IrDA IrLMP, 2835 defined, 12, 28 LM-IAS, 28, 3335 LM-MUX, 2833 as mandatory level, 28 structure, 29 tasks, 28 See also IrDA IR market, 16876 components manufacturers, 170 design guide, 169 mobile computing products, 171 products for desktop computers and peripheral devices, 172 service providers, 175 software products, 17374 IrOBEX, 191 defined, 191 structure, 192 IR-PLCP, 9798 header, 98 PLCP preamble, 97 PSDU, 98 IR-PMD, 99104 characteristics of signal used, 99 optical receivers, 1034 optical transmitters, 1023 pulse position modulation (PPM), 99102 IR transference protocol (IrTRAN-P), 12

231

JOLT, 176

class example, 34 defined, 33 elements, 33 frame C field structure, 35 frames format, 34 frames structure, 35 informational model, 34 structure illustration, 34 See also IrLMP Link management multiplexor (LM-MUX), 2833 access mode, 32 channels, 2829 command frame structure, 33 connection/disconnection, 32 data transfer, 32 defined, 29 discovery sniffing, 32 external connections example, 30 external interface, 2930 finite state machines (FSMs), 30, 31 flow control and, 29 frame format, 33 internal organization, 3032 internal structure illustration, 31 services, 3233 services access points (LSAPs), 29, 30 station control scheme, 32 status, 32 See also IrLMP Logical link control (LLC), 45 LR-HMPDU, 13940 defined, 139 structure, 140 See also HIPERLAN MAC PDUs (HMPDUs)

Law court scenarios, 156 LC-HMPDU, 13940 defined, 139 structure, 140 See also HIPERLAN MAC PDUs (HMPDUs) Link management information access service (LM-IAS), 28

MAC (HIPERLAN), 13146 address, 132 alias information base (AIB), 137, 144, 145 basis, 67 channel priority establishment, 138 data encyrption, 133

232

MAC (HIPERLAN) (continued) duplicate detection information base, 136 encryption-decryption scheme, 133 forwarder nodes, 132 functions, 13132 hello information base (HIB), 13637, 144, 145 HMPDUs, 139 IEEE 802.11 comparison, 6569 information database, 13537 MSAP, 111 neighbor information base (NIB), 136, 145, 146 nonforwarder nodes, 132 PDUs (HMPDUs), 111 priorities and traffic lifetime, 13739 p-saver information base, 136 p-savers, 13335, 134 p-supporter information base, 136 p-supporters, 135 route information base (RIB), 136, 145, 146 routing functions, 14246 source multipoint relay information base (SMRIB), 137, 144 topology information base (TIB), 137, 144, 145 MAC (IEEE 802.11), 4569 defined, 47 definition, 47 HIPERLAN comparison, 6569 service data units (MSDUs), 56 structure, 5466 for WLANs, 4769 Maximum adaptive defer threshold (MADT), 120 Medium access control (MAC) sublayer. See MAC (HIPERLAN); MAC (IEEE 802.11) Mobile broadband system (MBS), 5 Multicast support, 53


Multimedia Mobile Access Communications (MMAC), 189 Museum scenario, 15657 Neighbor information base (NIB), 136, 145, 146 Normalized residual lifetime (NRL), 138 Oplink Communications, 176 Optical receivers, 1034 Optical transmitters, 1023 Orthogonal frequency division multiplex (OFDM), 187, 188 Personal digital assistants (PDAs), 6 PHY (HIPERLAN), 111, 11321 burst characteristics, 115 channel access bursts, 118 channels, 114 data bursts, 11618 defer threshold establishment, 11921 delay spread, 116 goals, 113 HBR modulation, 116 introduction, 11314 LBR data burst, 116 LBR-HBR data burst, 11617 LBR modulation, 116 MADT calculation, 120 receiver characteristics, 11819 tasks, 114 transmission characteristics, 11416 transmitter/receiver classes compatibility, 119 transmitter types, 115 See also HIPERLAN standard PHY (IEEE 802.11), 45 DSSS, 8993 FH-SS, 8389 PLCP, 45, 87, 93 PMD, 45 transparency, 49 for WLANs (infrared systems), 96104

Index for WLANs (radio systems), 6996 PHY convergence protocol (PLCP), 45 DS-SS frame format, 93 FH-SS frame format, 8889 header, 98 IR, 9798 MAC interface, 96 preamble, 97 PHY medium dependent (PMD), 45 defined, 9697 IR, 99104 Point coordination function (PCF), 6366 defined, 56, 63 requirement, 63 timing diagram, 66 transmission procedure, 64 See also IEEE 802.11 standard Point-to-point protocol (PPP), 196 Power spectral density (PSD), 197 Public building scenarios, 15258 education, 15355 health care, 15556 high security, 15657 law court, 156 public transportation, 15758 requirements, 15253 See also Application scenarios Public-switched telephone network (PSTN), 3 Public transportation scenarios, 15758 Pulse position modulation (PPM), 17 4-PPM, 1819, 101 16-PPM, 101 data recovery, 99 duty cycle, 99 electrical spectrum of signal, 101 energy and time requirements, 101, 102 IR-PMD, 99102 QPSK, 79, 80 Route information base (RIB), 136, 145, 146 RTS-CTS procedure, 5863

233

Satellite UMTS (S-UMTS), 5 Second-generation (2G) wireless networks, 4 Self-collision, 52 Self-interference, 51 Slow frequency hopping (SFH), 72 Source multipoint relay information base (SMRIB), 137, 144 Spread spectrum systems, 7172 applications, 71 bandwidth, 71 complexity, 72 direct sequence (DSSS), 197200 techniques, 72 Spread spectrum techniques, 197200 benefits, 197 for diffuse links, 197200 performance, 197 STA flags, 16, 19, 21 Standards BRAN, 14748 evolution, 185, 18690 HIPERLAN, 10948 IEEE 802.11, 45105 IR, forthcoming, 190200 IrDA, 942 need for, 2 upcoming, 183209 STO flags, 16, 19, 21 SWAP, 207 characteristics, 208 defined, 207 See also HomeRF Symmetric DSL (SDSL), 3 Synchronized channel condition, 12730 acknowledgment, 130 contention phase, 127 defined, 127 elimination phase, 12829 prioritization phase, 127, 128 starting channel access cycle, 12728 transmission phase, 127 yielding phase, 12930 See also CAC

234

Synchronous data link control (SDLC), 23 TC-HMPDU, 14142, 143, 144 defined, 139 structure, 142 use of, 141 See also HIPERLAN MAC PDUs (HMPDUs) Technologies/products, 16176 IR market, 16876 RF market, 16168 See also Application scenarios Telephony, 185 Terminals DECT, 202, 203 design tendencies, 7 performance, 6 Terrestrial UMTS (T-UMTS), 5 Throughput, 48 Time division multiple access (TDMA), 45 TinyTP, 12, 3537 client services, 35 defined, 35 flow control, 3637 frames format, 36 protocol data units (TTP-PDUs), 36 services access point (TTPSAP), 35 See also IrDA Topology information base (TIB), 137, 144, 145 Universal mobile telecommunication system (UMTS), 5 University of New Hampshire Interoperability Lab (UNH-IOL), 168 Up-down collision, 52


Very high-speed DSL (VDSL), 3 Voltage controlled oscillator (VCO), 80 Wireless adaptive network device (WAND), 7 Wireless ATM (WATM), 4, 2067 Wireless Ethernet Compatibility Alliance (WECA), 162, 16465 current sponsors, 16465 defined, 164 Web site, 165 Wireless LANs advantages, 158 application scenarios, 15181 connectivity types, 55 future trends, 37 interconnection for, 19597 introduction to, 12 MAC protocol for, 4769 node support, 50 standardization need, 2 Wireless local loop (WLL), 6 WLAN Association (WLANA), 105, 161, 16264 affiliate member program, 163 affiliate members, 163 defined, 161 sponsor companies, 162 Web site, 162, 164 WLAN Interoperability Forum (WLIF), 162, 16566 aim, 166 defined, 162 members, 165 products certified by, 166

Wireless Lan Standards and Applications (Artech House Telecommunications Library)

Wireless Lan Standards and Applications (Artech House Telecommunications Library)

Telecommunications Technology Handbook (Artech House Telecommunications Library)

Chinese Telecommunications Policy (Artech House Telecommunications Library)

Wireless Internet Telecommunications (Artech House Mobile Communications)

Voice over 802.11 (Artech House Telecommunications Library)

Voice over 802.11 (Artech House Telecommunications Library)

Understanding Cellular Radio (Artech House Telecommunications Library)

World-Class Telecommunications Service Development (Artech House Telecommunications Library)

Customer-Centered: Telecommunications Services Marketing (Artech House Telecommunications Library)

Microwaves and Wireless Simplified (Artech House Antennas and Propagation Library)

Wireless Sensor Networks (Artech House Mems and Sensors Library)

Wireless LAN

Internet e-mail Protocols, Standards and Implementation (Artech House Telecommunications Library)

Digital Modulation Techniques, Second Edition (Artech House Telecommunications Library)

Open Source Software Law (Artech House Telecommunications Library)

Next Generation Intelligent Networks (Artech House Telecommunications Library)

Introduction to Error-Correcting Codes (Artech House Telecommunications Library)

Litigating With Electronically Stored Information (Artech House Telecommunications Library)

Litigating With Electronically Stored Information (Artech House Telecommunications Library)

Communications Technology Guide for Business (Artech House Telecommunications Library)

Understanding Voice over IP Security (Artech House Telecommunications Library)

Systems Approach to Engineering Design (Artech House Telecommunications Library)

Sip: Understanding the Session Initiation Protocol (Artech House Telecommunications Library)

Digital Modulation Techniques, Second Edition (Artech House Telecommunications Library)

Understanding SIP Servlets 1.1 (Artech House Telecommunications)

Wireless Technician's Handbook 2nd Edition (Artech House Mobile Communications Library)

Wireless Technician's Handbook (Artech House Mobile Communications Library) 2nd edition

Wireless Technician's Handbook (Artech House Mobile Communications Library) 2nd edition

Understanding Modern Telecommunications and the Information Superhighway (Artech House Telecommunications Library)

Fundamentals and Applications of Microfluidics (Artech House Microelectromechanical Systems Library)

Wireless Lan Standards and Applications (Artech House Telecommunications Library)