The Wireless Internet Opportunity for Developing Countries · The wireless Internet opportunity for developing countries ... By Amir Alexander Hasson Vendor Directory ... Internet

The Wireless Internet Opportunityfor Developing Countries

Edited by The Wireless Internet Institute

cover 11/19/03 9:45 AM Page 1

The Wireless Internet Opportunity for

Developing Countries

Funding for this publication was provided by

infoDev Program of The World Bank

United Nations ICT Task Force

Wireless Internet Institute

For ongoing online updates to this book,please visit w2i.org.

1.BookFrontPages 11/19/03 9:50 AM Page 1

The Wireless Internet Opportunity for Developing Countries is co-published by the Information forDevelopment Program (infoDev) of the World Bank; the United Nations ICT Task Force; and theWireless Internet Institute (W2i), a Division of World Times, Inc., to build on the momentum set by theconference of the same name cohosted by the Task Force and W2i at UN Headquarters in New York onJune 26, 2003.

This reference document serves as a basis for seminars and roundtable discussions to be held at theWorld Summit on Information Society in Geneva in December 2003 and provides a multi-annual livingknowledge base with ongoing updates in both electronic and printed form.

Copyright © 2003 by World Times, Inc.

All rights reserved, including the right to reproduce this book or portions thereof in any form or by anymeans, electronic or mechanical, including photocopying, recording, or by any information storage andretrieval system, without permission in writing form the publisher. All inquiries should be addressed tothe Wireless Internet Institute, A Division of World Times, Inc., 225 Franklin St., 26th Flr., Boston, MA02110, USA: www.w2i.org.

First printed November 2003

The findings, interpretations, and conclusions expressed in this document are entirely those of theauthors and the Wireless Internet Institute, and should not be attributed either to the World Bank, its affil-iated organizations, or members of its Board of Executive Directors or the countries they represent; or tothe United Nations ICT Task Force, its members, or subsidiary bodies.

Funding for the conference was provided by:Intel Corporation

This book is distributed on the understanding that if legal or other expert assistance is required in anyparticular case, readers should not rely entirely on statements made in this book, but should seek the serv-ices of a competent professional. Neither the Wireless Internet Institute, the UN ICT Task Force nor theWorld Bank accepts responsibility for the consequences of actions taken by those who do not seek nec-essary advice form competent professionals on legal or other matters that require expert advice.

ISBN 0-9747607-0-6


Welcoming Message from the Secretary GeneralKofi A. Annan

AcknowledgementsPrefaceExecutive Summary

IInnttrroodduuccttiioonnThe wireless Internet opportunity for developing countriesBy Mohsen Khalil

TTeecchhnnoollooggyy PPrriimmeerrOn spectrums and standards, architecture and access pointsBy Kevin C. Kahn

RReegguullaattoorryy EEnnvviirroonnmmeennttA global standard meets local policies and politicsBy Axel Leblois

Voice-over IP in potentiaFCC regulates the technologyMicrosoft chief reports on WRC-03

OOppppoorrttuunniittiieess aanndd RRooaaddbblloocckkssToward a broadband village area networkBy Sandy Pentland

For the poor, connectivity means economic opportunityBy Iqbal Z. Quadir

Wireless Internet ventures in developing countriesBy Theodore H. Schell

Contentsv

viixi

xiii

1

7

17

25

27

31

37

192021

CCaassee SSttuuddyy CCoommppeennddiiuummLeft to their own devices: Twelve developing-world case studiesBy Noopur Jhunjhunwala and Peter Orne

RReemmoottee RReeggiioonnssRising to the challege3Linking Everest, Khumbu Region, NepalRainforest IP3Xixuaú-Xipariná Ecological Reserve, Jauaperí River, Brazil

43

47


WWiirreelleessss IInntteerrnneett SSeerrvviiccee PPrroovviiddeerrssSlow incubation success3Africa-Online, Blantyre, MalawiMaking it at the margins3Baja Wireless, Ensenada, MexicoIt takes a hemisphere3E-Link Americas WISP on the edge3UniNet Communications, Cape Town, South Africa

SShhaarreedd AAcccceessssScaling Africa’s ivory towers3Partners in Education, Saint-Louis, Senegal; Njoro, KenyaVillage replication3Sustainable Access in Rural India, Tamil Nadu State, India

AAddaappttiivvee TTeecchhnnoollooggiieessOff the map3DakNet, Karnataka State, IndiaBeyond ‘off the shelf ’3Jhai Remote IT Village Project, Vientiane, Laos

RReebbuuiillddiinngg NNaattiioonnssWi-Fi Wild West3Internet Project Kosovo, Pristina, KosovoPicking up the PCs3UN Development Program, Kabul, Afghanistan

53

57

59

63

69

73

79

83

89

93

GGuuiiddeelliinneess ttoo CCoouunnttrriieess

VVeennddoorr RReessoouurrcceessUnderstanding the wireless Internet value chainBy Amir Alexander Hasson

VVeennddoorr DDiirreeccttoorryy

RReeccoommmmeennddeedd RReeaaddiinnggss

GGlloossssaarryy

IInnddeexx

99

107

111

117

119

125


v

It gives me great pleasure to send my greetings to the Wireless Internet Institute, theUnited Nations Information and Communication Technologies Task Force, and allother organizers and participants in this conference on wireless Internet opportunitiesfor developing countries.

As you know well, the swift emergence of a global “information society” is changing theway people live, learn, work, and relate. An explosion in the free flow of information andideas has brought knowledge and its myriad applications to many millions of people, cre-ating new choices and opportunities in some of the most vital realms of human endeavour.

Yet too many of the world’s people remain untouched by this revolution. A “digitaldivide” threatens to exacerbate already-wide gaps between rich and poor, within andamong countries. The stakes are high indeed. Timely access to news and informationcan promote trade, education, employment, health, and wealth. One of the hallmarks ofthe information society—openness—is a crucial ingredient of democracy and good gov-ernance. Information and knowledge are also at the heart of efforts to strengthen toler-ance, mutual understanding, and respect for diversity.

Wireless technologies have a key role to play everywhere, but especially in develop-ing countries and countries with economies in transition. With considerable speed andwithout enormous investments, Wi-Fi can facilitate access to knowledge and informa-tion, for example by making use of unlicensed radio spectrum to deliver cheap and fastInternet access. Indeed, it is precisely in places where no infrastructure exists that Wi-Fi can be particularly effective, helping countries to leapfrog generations of telecommu-nications technology and infrastructure and empower their people.

For this to happen, however, we urgently need to reach a clear understanding of Wi-Fi’s development potential, identify the obstacles, and develop a realistic plan of actionthat would bring together all stakeholders—governments, the private sector, civil socie-ty—in a coherent, synergistic, and sustainable endeavour. This conference offers a valu-able forum for doing just that. Your recommendations will also help guide the work ofthe UN Information and Communication Technologies Task Force. I wish you everysuccess and look forward to working with you on one of the leading development chal-lenges of our times.

U N I T E D N A T I O N S N A T I O N S U N I E S

THE SECRETARY-GENERAL

WELCOMING MESSAGE TO “THE WIRELESS INTERNETOPPORTUNITY FOR DEVELOPING COUNTRIES” CONFERENCE

New York, 26 June 2003

Kofi A. Annan



vii

CONTRIBUTING AUTHORS

Kofi A. Annan, Secretary General, United NationsAmir Alexander Hasson, Founder, First Mile SolutionsNoopur Jhunjhunwala, Analyst, Wireless Internet InstituteKevin C. Kahn, Director, Communications Technology Lab, Intel CorporationMohsen Khalil, Director, Global Information and Communication Technologies, The World BankAxel Leblois, Co-Founder, Wireless Internet Institute; President, World Times, Inc.Peter Orne, Editor, The WorldPaper, World Times, Inc.Alex (Sandy) Pentland, Founder, Digital Nations Consortium, MIT Media LabIqbal Z. Quadir, Lecturer, Harvard’s Kennedy School of Government; Founder, GrameenPhoneTheodore H. Schell, Founder, Cometa Networks

CONFERENCE STEERING COMMITTEE

Daniel Aghion, Executive Director, Wireless Internet InstituteShiv Bakhshi, Sr. Research Analyst, Global Wireless Infrastructure, International Data CorporationScott Borg, Research Fellow, CDS Tuck School of BusinessJeff Giesea, Publisher, FierceWirelessBrian Grimm, Director, Marketing, Wi-Fi AllianceYasuhiko Kawasumi, Chairman, New Technologies for Rural Applications Group, ITUMohsen Khalil, Director, Global Information and Communication Technologies, World BankSarbuland Khan, Director, Division for ECOSOC Support and Coordination, United NationsAxel Leblois, Co-Founder, Wireless Internet Institute; President, World Times, Inc.Amolo Ng’weno, Founder, Africa OnlineJeff Orr, Director, Marketing, WiMax ForumAlexander (Sandy) Pentland, Founder, Digital Nations Consortium, MIT Media LabBimal Sareen, Former CEO, Media Lab Asia; International ConsultantAlan Scrime, Chief, Policy and Rules, Office of Engineering and Technology, FCCKhaldoon Tabaza, Founder, Arab AdvisorSriram Viswanathan, Managing Director, Strategic Investments, Intel CapitalLinda Wellstein, Partner, Wilkinson Barker & Knauer, Llp.

AcknowledgementsThe publishers—the Information for Development Program (infoDev),

United Nations ICT Task Force, and the Wireless Internet Institute (W2i)—wish to acknowledge the following individuals and organizations for their

contributions to “The Wireless Internet Opportunity for Developing Countries”conference, research, and publication program.


THE WIRELESS INTERNET OPPORTUNITIES FOR DEVELOPING COUNTRIESviii

CONFERENCE FACULTY

Talal Abu-Ghazaleh, Vice Chairman UN ICT Task Force; CEO, TAGILisa Agard, Executive Vice President, Cable & Wireless Daniel Aghion, Executive Director, Wireless Internet InstituteShiv Bakhshi, Sr. Research Analyst, Global Wireless Infrastructure,

International Data CorporationScott Borg, Research Fellow, CDS Tuck School of BusinessBill Carney, Board Member, Wi-Fi AllianceAmir Dossal, Executive Director, United Nations for International PartnershipsPat Gelsinger, Chief Technology Officer, Intel CorporationVic Hayes, Sr. Scientist, Agere SystemsMohsen Khalil, Director, Global Information and Communication Technologies,

The World BankSarbuland Khan, Director, Division for ECOSOC Support and Coordination, United NationsEd Malloy, USAIDAlex (Sandy) Pentland, Founder, Digital Nations Consortium, MIT Media LabRobert Pepper, Chief, Plans and Policy, FCCIqbal Quadir, Lecturer, Harvard’s Kennedy School of Government; Founder, GrameenPhoneBimal Sareen, Former CEO, Media Lab Asia; International ConsultantTheodore H. Schell, Founder, Cometa NetworksMohammad Shakoury, Board Member,

WiMax ForumJabari Simama, Executive Director, Atlanta Mayor’s Office of Community TechnologyFrancisco Vieira, Operational Specialist, IT4D, Inter-American Development BankSriram Viswanathan, Managing Director, Strategic Investments, Intel CapitalLinda Wellstein, Partner, Wilkinson Barker & Knauer, Llp

FIELD PRACTITIONERS

Dileep Agrawal, CEO and Co-Founder, WorldLink ISP Michael Best, Board Member, SARI ProjectChristopher Clark, Co-Founder, Amazon AssociationRobert Freeling, Executive Director, Solar Electric Light FundAmir Alexander Hasson, Founder, First Mile SolutionsDavid Jarvis, Founder, UniNet CommunicationsRafiqullah Kakar, National Project Coordinator, ICT4D, UNDP-AfghanistanPaul Meyer, Founder, Internet Project Kosovo; Co-Founder, CEO, VoxivaMargery O’Donnell, Project Administrator, John W. McCormack Graduate School of

Policy Studies, UMass BostonPaul Shaw, CEO, Africa-OnlineSharron Tate, Co-Founder and Director of Operations, Baja WirelessLee Thorn, Chairman and Founder, Jhai FoundationRandy Zadra, Managing Director, Institute for Connectivity of the Americas


ACKNOWLEDGEMENTS ix

CONFERENCE SHOWCASE PRESENTERS

Intel CorporationNomadixRadiant Networks, Plc.Telesym, Inc.Tropos NetworksVivato, Inc.

UNITED NATIONS ICT TASK FORCE

Sarbuland Khan, Director, Division for ECOSOC Support and Coordination, United NationsSergei Kambalov, Deputy Executive CoordinatorSamuel Danofsky, Program ManagerMaria Lehtinen, Special Assistant to the DirectorEnrica Murmura, Media and Outreach OfficerDaniella Giacomelli, Program ManagerJulianne Lee, Senior Policy Advisor to the ChairmanLinda Sanchez, Administrative Assistant

INFODEV

Bruno Lanvin, Program ManagerJacqueline Dubow, Program Coordinator

EDITORIAL ADVISORY BOARD

Bruno Lanvin, Program Manager, infoDevJacqueline Dubow, Project Coordinator, infoDevSarbuland Khan, Director, Division for ECOSOC Support and Coordination, United NationsSergei Kambalov, Deputy Executive Coordinator, UN ICT Task ForceTheodore H. Schell, Founder, Cometa NetworksPeter Orne, Editorial Director, W2i

W2i EDITORIAL AND PRODUCTION TEAM

Daniel Aghion, Project ManagerKeith Frazier, Art DirectorNoopur Jhunjhunwala, AnalystAxel Leblois, Contributing EditorChristine Leblois, Project AdministratorPeter Orne, Editorial Director



xi

The promises of wireless Internet technologies have generated much interest onthe part of the international-development community. While in developed nationsthese technologies have primarily been associated with mobility applications and

local area networking in homes and offices, their most intriguing application in devel-oping nations is the deployment of low-cost broadband Internet infrastructure and last-mile distribution.

The rationale for such interest is simple in theory: The digital divide cannot beresolved any time soon because of the prohibitive cost of deploying conventional wiredinfrastructure in developing countries. Wireless Internet, however, has the potential tosolve this bottleneck, as the collection of articles and case studies in this volume demon-strates. And leading IT vendors do confirm that wireless Internet should be the mostpromising accelerator of technology adoption in developing nations. In support of thesehopes, technology is readily available and more accessible every day thanks to stan-dardization and fast-declining costs.

So, why should this topic become central to the World Summit on InformationSociety initiative?

First, wireless Internet may be a very effective and inexpensive connectivity tool, but itdoes not carry any magic in itself. It can only be successfully deployed as demand for con-nectivity and bandwidth emerges in support of relevant applications for the populationsserved. These may be supporting e-government, e-education, e-health, e-business or e-agriculture applications. But those are not easily implemented in the developing world.

Demand aggregation for wireless Internet connectivity around applications that makesense in support of wireless infrastructure investment is the first important challengethat the UN ICT Task Force, infoDev, and the Wireless Internet Institute wanted toexplore and document. The authors of this compendium have investigated dozens offield experiments around the world and selected several that exemplify some of theinnovative approaches to this challenge. They do suggest that wireless Internet canindeed be sustainably and in some cases profitably deployed in support of economicand social development objectives in developing countries.

One common characteristic of these case studies is their unconventional, often grass-roots origin. Entrepreneurs from the private, public, or not-for-profit sectors have inde-pendently developed original deployment models pointing to potential solutions for thedeveloping world. Most, however, must confront serious challenges that are nontechni-cal in nature and associated with legacy regulations, administrative obstacles, and theopposition of incumbent telecommunications operators.

This book attempts to clarify the key issues that need attention on the part of gov-ernments, international development agencies, and nongovernmental organizationswhose goals are to bridge the digital divide. It is a first step only and will be comple-mented by further updates and analysis. Documenting this worldwide phenomenon,sharing best practices and experiences, and promoting adequate policies can help toaccelerate the resolution of the many obstacles that constrain the expansion of one of themost promising information-technology tools of this century.w

Preface



xiii

Executive summary

T he Wireless Internet Opportunity forDeveloping Countries examines the emer-gence and promise of proven and inexpensive

technologies to bridge the connectivity gap at theroot of the digital divide. It brings together the per-spectives of thought leaders from the fields ofsocial and economic development, informationand communication technology, government, reg-ulation and international standards; and the exam-ples, collected in a case-study compendium, ofwireless Internet field practitioners at workthroughout the developing world.

When affordable and relevant, information andcommunications technologies have been adoptedat a fast pace within developing countries,leapfrogging traditional infrastructure. Followingthe fast growth in developing countries of cellulartelephones over fixed lines—in Africa, the ratio isapproaching 2 to 1—Internet applications sup-porting telecommunications as well as economic,education, and social services activities have prolif-erated rapidly. In Brazil, for example, 90 to 95 per-cent of tax returns are now filed online, a fargreater percentage than in North America.Connectivity and the information access it bringsunleashes human capital and increase productivityand knowledge sharing in underserved areaswhere it has been most constrained.

New wireless Internet technologies are ready tofulfill this promise, supported by universallyaccepted standards set by the Institute of Electricaland Electronics Engineers (IEEE) for both wire-less local area networking (the 802.11 standardsknown as “Wi-Fi”) and long-distance point-to-point connectivity (the 802.16 standards known asWiMax). A multitude of devices, software, andservices on the market are designed to inter-oper-ate with unified protocols in the frequency spec-trums defined by these standards.

In recent years, the definitions of unlicensedspectrum have normalized around two major setsof frequencies: the Instrument, Scientific, and

Medical band (ISM) at 2.4 Ghz; and a newer allo-cation in the 5-Ghz to 6-Ghz range adopted by theWorld Radiocommunication Conference (WRC)in June 2003. More and more devices are reachingthe market with chip sets that will allow the instantdetection and connectivity to local hotspots wherethey are available.

Developing countries, however, are presentedwith a number of technological trade-offs. Forexample, in the absence of wired infrastructure,wireless point-to-point connection can be usedvery effectively to provide inexpensive broad-band access. But online interactive applicationsand voice-over IP require as little “latency” aspossible. As the number of hops grows along aconnection, so does latency. And so-called multi-ple-hops configurations will often end up in acommon link to a high-performance fiber back-bone, presenting yet another potential bottle-neck. Developing countries will need to establisha minimum level of performance based on thecapacities of local infrastructure.

Fast declining costs promise to make wirelessbroadband more and more affordable, especially asit is embedded in more devices, including “intelli-gent phones” and other hardware that is able toautomatically shift from a local 802.xx connectionwith voice-over IP to a regular cellular service.Perhaps more important for developing countriesis new “smart” technology under developmentthat will help to optimize spectrum use by auto-matically limiting interference, solving a potential-ly difficult issue in dense areas.

Equally intriguing for developing countries isnew technology allowing users to automaticallyrelay radio signals, which creates a “mesh net-work” of wireless connections that could develop alife of its own, reducing the number of requiredbase stations. This technology lowers the cost ofinfrastructure while increasing the cost to usersonly marginally, and provides welcome connectivi-ty redundancy in dense areas.


THE WIRELESS INTERNET OPPORTUNITY FOR DEVELOPING COUNTRIESxiv

Because many developing countries areadvanced in the adoption of wireless cellulartelephony, the topic of the future convergencebetween IEEE 802.xx standards and cellular stan-dards of the International TelecommunicationUnion (ITU) is obviously important. The 802.16standard, or WiMax, was defined for broadbanddistribution to fixed points on reasonably large-channel allocations, whereas mobile telephonystandards are designed to transport voice on nar-row bands. Both, however, are evolving towardeach other’s position: 802.16e supports someform of nomadic applications while 3G, with ahigh level of mobility, offers mid-bandwidth dataservices. 4G and 802.20 may still be at the con-ceptual level, but developing countries will need toremain well informed in order to optimize existinginfrastructure, given that these technologies willcontinue to evolve.

Amid the fast pace of technological advance,regulatory authorities are faced with a number ofissues. Defining international unlicensed bands insupport of wireless Internet and other applicationshas been long and arduous, but the process hascome to fruition thanks to the active role of severalkey constituents including the US FederalCommunications Commission, the EuropeanUnion, the ITU, and the IT industry. The 2003World Radio Conference decision to allocate the2.4-Ghz and 5-Ghz bands to license-exempt appli-cations, as voted on by participating countries, wasa landmark for future universal adoption. Today,41 percent of developing countries have regula-tions supporting license-exempt bands, comparedwith 96 percent among developed countries. Andmany developed countries, including the UnitedStates and the European Union, have regulationsthat now support the commercial use of license-exempt spectrum.

One of the issues among developing countries isthe delicate transition from telecom monopolies tofree competition and a deregulated environment.Legacy infrastructures, old regulations, lack ofcompetition, the wish to preserve monopolies’cash flows, and political issues related to controlover information all contribute to the slow the

adoption of new regulations supporting license-exempt bands and their free use for commercialpurposes.

Even where license-exempt bands are author-ized, additional regulations still limit its use. Somecountries impose license fees on individual equip-ment for hotspots and access devices. In theabsence of a clear regulatory framework support-ing competition in telecommunications, access toan Internet backbone may be made either impossi-ble or cost-prohibitive by incumbent telecoms orthrough licensing by regulatory authorities.

In this context, voice-over IP is perceived aspotentially disruptive and has triggered forcefulopposition from the governments of many devel-oping countries. For the purpose of bringingwireless Internet to underserved areas, however,VoIP may play a very positive role in providinginexpensive telephone access and helping localwireless Internet service providers to benefit fromhigher traffic.

The IT industry views the June 2003 decisionof the World Radio Conference as a long-term pos-itive impact, opening an unlimited field of oppor-tunities—a win-win on a worldwide basis for serv-ice providers, hardware and software companies,and most importantly customers.

Today’s wireless economics are already com-pelling: Wireless local loops are about one thirdthe cost of copper or fiber land-line service, whilepacket-based broadband computer networks costone ninth of land-line service. Ease of set-up, use,and maintenance are affordable for both users andproviders. Tests in rural settings show that a $30wireless PC card can provide good connectivity upto a half-kilometer radius with line-of-sight and upto 20 kilometers with antennas and repeaters. AndWi-Fi access points can be purchased for $80.

Developing basic asynchronous services may bethe best starting point for rural connectivity,including voice messaging, e-mail, and batchdownloading.

MIT Media Lab researchers have developedand demonstrated a combination of wireless tech-nology and asynchronous delivery that brings thecost of asynchronous Internet service by two


EXECUTIVE SUMMARY xv

orders of magnitude below land-line expenses.These observations suggest that developing-coun-try governments should adopt regulatory frame-works that support independent wireless Internetservice providers and that do not restrict voice-over IP applications.

The inspiring mobile-telephony success storyof GrameenPhone in rural Bangladesh shows howentrepreneurial villagers can be recruited to adopta technology that provides them greater connectiv-ity and therefore economic opportunity on a wholenew scale. GrameenPhone has built the largest cel-lular network in the country, in numbers of sub-scribers, with investments exceeding $300 millionand a subscriber base of more than one million. Itsrural program is available in more than 35,000 vil-lages, providing telephone access to millions whilefostering a generation of village micro-entrepre-neurs. GrameenPhone is an example of how wide-spread connectivity can dramatically unleashhuman potential and increase productivity. Themodel may be applicable with limited variations towireless Internet services that generate localemployment and ownership.

Experienced industry leaders, however, point tothe many obstacles encountered by start-up wirelessInternet service providers in developing countries.These include operating in areas with low dispos-able income, low initial demand for wireless servic-es because of a small PC base, difficulty aggregatinglocal demand, and limited financing options. A pri-mary question, therefore, is how the revenue/costequation is solved. Moreover, despite favorable reg-ulations and stated support for ISPs, incumbenttelecoms may create significant hurdles with preda-tory pricing, untenably high bandwidth rates, denialof interconnections for long periods, delays inrepairing leased lines, or failure to meet contractuallevels of services. Similar issues are encountered indeveloped countries, but in developing countriesISPs have less economic and legal leverage.

Case studies of field deployments of wirelessInternet networks raise the many issues discussedabove in a multiplicity of operating environments.In the run-up to the UN ICT Task Force andWireless Internet Institute conference held at UN

Headquarters on June 26, 2003, dozens of casestudies were gathered, of which twelve are collect-ed here to showcase some of the variation in oper-ating environments, sizes, business models, andpurpose.

Whether motivated primarily by developmentobjectives, for profit, or by the challenges ofdeployment, the field practitioners in these pio-neering case studies share the goal of bridgingthe digital divide using broadband wireless tech-nology. Each found a unique solution for over-coming roadblocks.

The case studies analyzed for this compendiumappear in five categories revealing core aspects ofwireless broadband deployments in developingcountries:3Remote Regions provide the harshest environ-mental test of the technology and of the implemen-tation models (pp. 43-50).3Wireless Internet Service Providers are prima-rily driven by profit but can also contribute toregional development (pp. 53-66).3Shared Access provides connectivity to regionsthat cannot afford individual access (pp. 69-76).3Adaptive Technologies rethink off-the-shelfsolutions to overcome local limitations (pp. 69-85).3Rebuilding Nations can draw from the modelsabove to rebuild a communications infrastructureand jumpstart social and economic development(pp. 89-95).

While examining these case studies, analysts ofthe Wireless Internet Institute confirmed keyanalysis offered by experts on a more “macro”level:3In most areas, there is simply no wired alterna-tive available either for geographic and social envi-ronment, economic, or maintenance reasons.3Most case studies show early adopters innovat-ing as they set-up their network.3Inter-connection equipment and service to abackbone are a major expense.3In remote areas, power supply remains a difficultissue.3Explaining the benefit of broadband and gettingthe “buy-in” of local communities is crucial and


THE WIRELESS INTERNET OPPORTUNITY FOR DEVELOPING COUNTRIESxvi

demand aggregation is often necessary, includingthrough community centers or kiosk models.3When people connect to the Web, they want tocommunicate, and e-mail and voice and video mailare the most popular uses of wireless broadbandnetworks in the developing world.3The Internet’s biggest advantage over other com-munication technologies is that it can provide anarray of solutions and products for e-governance,e-health, e-education, and e-commerce.3Incumbent telecom operators in the developingworld guard their terrain carefully. Rural areas aremarginal markets for incumbents, so entrepre-neurs may face few initial barriers, whereas incities, the entrepreneur is likely to face the incum-bent and regulatory authority head on.3Different responses to these entities, however,present varying risks. Some operators choose todirectly confront telecom operators and authori-ties, lobbying against them domestically and inter-nationally, which can lead to confiscations, projecttermination, and even criminal charges. Others

bypass the regulatory environment altogetherwhen the authority is not too aggressive. Still oth-ers engage in the bureaucratic process, seekingrequired permissions and even trying to formalliances with incumbents.3Authorities may be more convinced by projectswith a social-development outcome.

Local governments are encouraged to develop agreater awareness of the potential benefits of wire-less Internet connectivity, to foster local initiatives,and carefully adjust their regulatory frameworks,taking into account worldwide standards, barriersto competition, local Internet backbone accessconditions, and the monitoring of early wirelessInternet experiments.

Entrepreneurs and field practitioners lookingfor resources or simply seeking assessments ofindustry and technology trends will find an analy-sis of the wireless Internet value chain followed bya directory of firms and technologies. For ongoingonline updates to this book, visit and contact theWireless Internet Institute at w2i.org.w


INTRODUCTION 1

By Mohsen Khalil

When it comes to wireless technology, Ido not believe we’re going to see theusual classical behavior, where there will

be a technology transfer from the industrializedworld to the developing world. There may actu-ally be a reverse flow of innovative solutions, andthe developing world might in fact dictate thestandards and bring many solutions to the glob-al market.

At the World Bank, our interest is in achievingconnectivity access and trying to devise solutionsin the developing world, regardless of what thetechnology platform is. If we see an opportunity ina specific technology, we’ll take it, but that is notthe object itself. We truly believe there is a directlink between access and poverty reduction. In thedeveloping world, access presents an unprece-dented opportunity for people who have been dis-connected from the global economy.

One project in Bangladesh, which was partlyfinanced by IFC, has been very successful.GrameenPhone has surpassed the incumbent andworked well in an inefficient, regulated environ-ment to bring cellular telephony to rural areas.Women in the villages did this by providing com-munity solutions and services, and many on aver-age earned double their income per capita.

Another example of how access providesopportunity is the provision of online informationto farmers, so they can watch for lower costs and

improve their profit margins. And yet anotherexample is Novica, the largest portal enabling arti-sans in rural areas to sell their products abroad—for example, to department stores in the UnitedStates. The World Bank has just provided themwith an equity investment.

In a broader way, IT can provide a much moreefficient platform for the delivery of social servicesand health and education. One of the imperativesfor creating a proper investment climate andimproved economic growth is improved govern-ment and transparency. In Brazil today, for exam-ple, 90 to 95 percent of all tax returns are doneonline, which, in comparison, is much higher thanin the United States. IT also helps people, fromNGOs to interest groups, to be able to connect tothe global community and voice their rights.

Unlike the industrial or agricultural revolutions,where you needed huge amounts of capital tooperate, once you provide information access, youhave the power to unleash the potential of humancapital no matter where you are.

Perhaps the Internet bubble hasn’t served uswell, because people thought about this revolutionin the context of stock markets. I do not believe it’sover, however. It’s a dormant volcano that willcome back again. It allows an opportunity to enterthe global market, it allows improved efficiencythroughout the value chain, and ultimately it’sgoing to have an impact on the political texture ofany society. For example, coverage of the war inIraq wouldn’t have been the same; you couldn’thave access to information. None of what happensin the world today would have been the same twen-ty or thirty years ago.w

The wireless Internet opportunityfor developing countries

Mohsen Khalil is Director of the Global Informationand Communication Technologies Division of the TheWorld Bank.

1.Pages1_End 11/19/03 10:17 AM Page 1

THE WIRELESS INTERNET OPPORTUNITY FOR DEVELOPING COUNTRIES2

Global connectivity is on the rise, thanks largely to cellular technology. In 2001, the number of fixed lines andthe number of cellular lines crossed. Since then, we have had more mobile phones than fixed lines. The growthrate of mobiles is significantly higher than fixed lines. The blue dots on the map show all the countries andregions where there are more mobile phones than fixed lines. Today, there is more mobile telephony in devel-oping countries than in industrialized countries.

Although it has mostly been concentrated in the industrialized world, there has been an explosion in Internetgrowth. You can see almost exponential growth in Internet users, and in almost every single country today thereis some form of Internet connectivity.

GLOBAL CONNECTIVITY RISING

INTERNET GROWTH SLOW BUT SURE

1.Pages1_End 11/19/03 10:17 AM Page 2

INTRODUCTION 3

The access gap is narrowing, but the information gap is widening. In terms of telephone connectivity, the gapbetween the high-income countries and the middle-income countries is narrowing. But if you measure the gapby connectivity of Internet hosts, the gap is widening in a very serious way.

The Internet access gap is widening, which is very concerning. Although the gap has been narrowing amongregions of the developing world, access has been concentrated in the 10 largest economies. There are seriousgaps between different regions. For example, the gap between East Asia and South Asia is huge. There arehuge gaps between rural and urban (improving somewhat) and between the educated versus not educated.

BRIDGING THE DIVIDE

INTERNET ACCESS GAP ENDURES

1.Pages1_End 11/19/03 10:17 AM Page 3


The growth rate of mobile phones is expected to be significantly higher than fixed-line growth. We’re approach-ing a point where for every fixed line in Africa, we have two cellular lines, at least. If you compare this withNorth America, it’s the opposite.

Among the Least Developed Countries, we have experienced a 50- to 70-percent compounded rate of growthfor mobile phones on an annual basis. That’s great news. Much more was done in the past 10 years than overthe previous 50 years, and much more can be done, but the challenge is how to do it. On the Internet, thoughthe growth rate is high, we’re starting from an extremely low base.

THE TRIUMPH OF MOBILE

LDC’S ARE MAKING GAINS

SOURCE: PYRAMID

1.Pages1_End 11/19/03 10:17 AM Page 4

TECHNOLOGY PRIMER

1.Pages1_End 11/19/03 10:17 AM Page 5

1.Pages1_End 11/19/03 10:17 AM Page 6

TECHNOLOGY PRIMER 7

By Kevin C. Kahn

W ireless technology is revolutionizingthe way in which PCs, PDAs, and othersimilar devices access the Internet at high

speeds. This is happening both in the last 100meters where it provides a mobile alternative toEthernet, and in the longer-reach access networkswhere it complements technologies like DSL andcable modems. This primer provides a technicaloverview of the various wireless technologies cur-rently or soon to be in use, how and in what spec-trum they operate, and where they each fit in alarger Internet-access architecture.

Basic Spectrum Issues

The radio spectrum is used by a wide variety ofusers ranging from consumer radio and television,to weather and aircraft radar, to data communica-tions. Ranges of frequencies are assigned to vari-ous uses based on history, technical properties ofthe various frequencies, and other considerations.Because radio waves do not respect national bor-ders, international agreements attempt to harmo-nize uses across multiple countries where conflictscould occur or where there is benefit using com-mon equipment. For this primer, we are interestedin the use of frequencies for short- to medium-range data communications.

To control how various bands are used, nationalregulations have commonly required licenses foroperators to be permitted to use particular frequen-

cies. Licensed bands, or radio channels, are mostimportant to relatively high-power and longer-rangeuses of radio where significant chance of interfer-ence between different radio users would exist ifcontrol is not exercised. For example, TV channelsare allocated so that multiple broadcasters do notoperate so near to one another that receivers wouldsee their reception garbled.

For some types of radio use, requiring licenses istoo cumbersome, so many national regulators havealso defined unlicensed or license-exempt bands.To maintain order in such bands, there are gener-ally rules limiting the power and other technicalproperties that a radio operating in the bands canuse. Because of the typical low power used, usersare expected to be able to use such bands eitherwithout mutual interference or by managing anyinterference among themselves without govern-mental legal help.

The most commonly discussed unlicensedband, available virtually worldwide today, is in thevicinity of 2.4 GHz. This band is often called theIndustrial, Scientific, Medical (ISM) band becauseits initial allocation was to allow radio emissions byvarious sorts of equipment. This is the band that isbeing used today for Wireless Local AreaNetworks (WLAN) according to the IEEE802.11b/g standards and branded by an industrygroup as Wi-Fi.

Another commonly discussed set of bands arein the space between 5 GHz and 6 GHz wherethe IEEE 802.11a standard is defined to operate.The unlicensed allocations in this band havebeen the subject of recent international harmo-nization efforts through the ITU at the 2003

On spectrums and standards,architecture and access points

Kevin C. Kahn is an Intel Fellow and Director of theCommunications and Interconnect Technology Lab,Corporate Technology Group, Intel Corporation.

1.Pages1_End 11/19/03 10:17 AM Page 7


World Radiocommunication Conference (WRC-03). (For more information, see www.itu.int/ITU-R/conferences/wrc/wrc-03/index.asp.)

The Technology Lineup

802.11

The major standard body for defining unlicensedwireless communications schemes is the IEEEwhere these standards are part of the 802 series ofstandards. The best known of these wireless stan-dards today is the 802.11 series which describeWLANs. These radio systems allow communica-tions within a relatively small area, typically withinabout 100 meters of a central access point (AP).

A typical use of 802.11 today would be to placeAPs strategically around an office, each connectedto the wired Internet, and thus allow any PCs orother data terminals in the area to get access to thenetwork. These APs may be deployed privatelyand secured so that only, for example, employeesof a business can use them, or they may bedeployed publicly in a common place such as anairport or train station where any users may usethem (either for free or for a service charge).

Any wireless data communications standardprimarily consists of two layers. The first is a defi-nition of a protocol that lets stations using theradio channel know when it is their turn to trans-mit and how actual data is packaged with informa-tion about who it is for and where it starts andstops in the signal. This layer is called the MediaAccess Control layer, or MAC. Below this layer is adefinition of how an actual radio signal is modulat-ed to encode information on it, which is called thephysical layer, or PHY.

For 802.11, the MAC is optimized for a set ofequally important systems as one typically finds ona local area network. A central AP manages the col-lection, giving each station that needs to sendinformation a turn to do so. Since the AP may notalways know that a remote system needs a slot,there are also periodic times when any station maytransmit without permission. Because multiple sta-tions may decide independently to do so at thesame time, this sort of transmission is subject to

collisions that require stations to waste timeretransmitting their information. The combinationof these managed and unmanaged periods, howev-er, creates a stable system that allows all the net-work users to fairly share the limited capacity ofthe channel.

In the 802.11 family, the main standards aredesignated 802.11b, 802.11a, and 802.11g. Thesediffer primarily in their definition of what PHY isused, though each expects to operate using about a20-MHz wide channel. The first version to bedeployed operated in the 2.4-GHz band and wasdesignated 802.11b. The 802.11b standarddefines operations at several different data rates,ranging from 1 Mbps up to 11 Mbps in each chan-nel. How much data you can successfully transmitthrough a channel is primarily related to howstrong the signal is compared with the backgroundnoise. This is similar to how well you can hearsomeone talking being a function of how loud theyspeak compared with how noisy the environmentis. When the signal-to-noise ratio is good, you canachieve the higher data rates. Note that the fartheraway from a transmitter a receiver is located, theweaker the received signal will be. Thus poor sig-nal-to-noise ratios can be the result of interferenceor simply that the source is distant.

The 802.11b PHY was designed to allowdevices to efficiently share spectrum and toimprove on earlier transmission schemes that hadinferior performance in the presence of radio inter-ferers (noise) and reflected signals (called multi-path). The 802.11b PHY uses a scheme calledDirect Sequence Spread Spectrum (DSSS) totransmit signals over the air. DSSS provides someprotection against interferers such as the radiationfrom microwave ovens or 2.4-GHz cordlessphones. 802.11b’s lower data rates (1 Mbps and 2Mbps) use a technique called Barker coding inconjunction with DSSS which can work in poorsignal conditions; whereas 802.11b’s higher datarates (5.5 Mbps and 11 Mbps) use an alternativetechnique called Complementary Code Keying(CCK) in conjunction with DSSS.

There are only a limited number of distinct 11bchannels available because typically this band is

1.Pages1_End 11/19/03 10:17 AM Page 8

TECHNOLOGY PRIMER 9

less than 100MHz wide and other radio alloca-tions on either side mean that it isn’t practical toexpand the band internationally. To make availablemore WLAN capacity, regulators have agreed toallocations between 5 and 6 GHz. 802.11a is a ver-sion of the standard designed to operate in thesebands, and it is now beginning to come on the mar-ket. This version of the standard also differs fromthe 802.11b standard by using more advancedPHY techniques that allow it to have a maximumdata rate of 54 Mbps. A more recent standardcalled 802.11g applies similar advanced PHYtechniques in the 2.4 GHz band to achieve opera-tion up to 54 Mbps. Looking past 802.11a and802.11g, work continues in the standards body tolook at even more advanced PHY techniques tocontinually improve the robustness and speed ofWLAN systems.

The 802.11 standards committee also definesother aspects of the standard in associated specifi-cation such as 802.11e which will describe quali-ty-of-service enhancements and 802.11i whichwill improve security protocols.

As 802.11 systems have become increasinglyavailable and useful, we are now seeing 802.11

capabilities being built into more consumerdevices. For example, all mobile PCs that useIntel® CentrinoTM Mobile Technology comeequipped to access the network using 802.11. Weare also beginning to see PDAs and even phonesthat use 802.11 to communicate.

While 802.11 standards are designed technical-ly for short-range operations, some users have cre-ated somewhat nonstandard equipment that iscapable of longer-range operation of up to multiplekilometers. The required modifications includespecial antennas, boosted power output, andchanges to detailed aspects of the data encoding toallow for the longer time taken by signals that needto travel such distances. A better and standardizedsolution to longer-range data communications canbe found in a parallel standard that has recentlybeen developed, IEEE 802.16.

802.16As with 802.11, 802.16 is actually a family of

standards, some completed and some still inprogress. This family also has an associated indus-try group, WiMax. The original 802.16 standarddefines a MAC suitable for an access system based

THE 802.16 STANDARD (WI-MAX) ADDRESSES BACKHAUL

1.Pages1_End 11/19/03 10:17 AM Page 9


at a central base station and serving multiple usersscattered over a relatively large area whose radiuscan be many kilometers. This version of the stan-dard also defines a particular physical layer (PHY)that is suited for use in bands between 10 and 66GHz. The standard is optimized for providing anaccess service to, for example, an entire buildingwhere multiple systems within the building can allbe attached to the single building transceiver.

It is helpful to think of 802.16 as an alternativeto a wire or fiber for delivering an Internet connec-tion to a site. Unlike 802.11, this standard does notnarrowly define the frequency bands that shouldbe used nor limit the channel width. It can there-fore accommodate whatever bands, licensed orunlicensed, that may be available to an operator.For example, with a 20-MHz channel available anda strong signal between the base station andremote site, the capacity can be as high as 96 Mbpswhile a 28-MHz channel in a similar situationcould achieve over 130-Mbps. On the other hand,remote sites at the fringe of reception of a base sta-tion might only be able to achieve a 32-Mbpscapacity. Of course, as a multipoint system expect-ed to serve many customers, these capacities mustbe shared across all the served users.

Another member of this standards family,802.16a is optimized for operation in frequenciesbetween 2 and 11 GHz. It is also yet more flexiblein channel width choices, including channels asnarrow as 1.75 MHz to allow it to be used whereonly small allocations are available. This version isattracting considerable commercial attention nowbecause this range covers a number of attractivebands found around the world.

First, the unlicensed bands described earlierfall into this range. 802.16a systems are long-range systems and therefore at first look do notappear attractive for use in unlicensed bandswhere interference among competing operatorscould become a problem. However, particularly inrural and developing markets it is likely that therewill be sufficient unlicensed spectrum and littleenough competition for it that operators may findits use quite reasonable.

There are also commonly available bands at 2.5

GHz and 3.4 GHz in various countries that arelicensable for use with an access data service.802.16a looks to be a good choice for these sys-tems as well. An operator licensed to have exclu-sive use of part of one of these bands could offerbroadband wireless access in more densely popu-lated urban or suburban areas without interfer-ence concerns.

One key issue for wireless access systems suchas 802.16 is whether they require line-of-sight, orLOS, between the receiver and the base station orwhether near LOS or non-LOS is sufficient.Ideally, LOS would not be required, but the realityis that radio waves are always attenuated whenpassing through obstacles so that non-LOS per-formance will always be poorer than LOS per-formance. For good performance, the best designwill be an end station that mounts at least theantenna on the outside of the building facing gen-erally in the direction of the base station. Whilepossibly complicating installation slightly, this willassure the best range and performance for an802.16 deployment.

3G and beyondIt is worth briefly relating these data communica-

tions standards with the evolution of conventionalcellular telephony as it adds data communicationsservices. 802.16 starts from the premise of deliver-ing broadband data to fixed points. For example, itgenerally assumes a reasonable wide channel alloca-tion. To this it is adding mobility capabilities via802.16e, which will allow it to support at least anomadic model; that is, one where an end stationdoesn’t move much while operating but may movearound between sessions. It has also been addingsupport for narrow channels. Higher degrees ofmobility will also require considerably more sup-port for handing active connections off betweenbase stations. On this infrastructure, one can thenthink about running Voice-over Internet packets, orVoIP, to provide standard telephony service.

3G cellular systems start with the premise ofdelivering highly mobile voice services andincreasing narrow- to mid-bandwidth data servic-es. Their infrastructure is optimized for high

1.Pages1_End 11/19/03 10:17 AM Page 10

TECHNOLOGY PRIMER 11

mobility including high-speed handoff. Data serv-ices are carried over a somewhat more complextechnical structure designed for these needs.

For the immediate future, what infrastructure todeploy will be determined by previous investmentsin infrastructure (e.g., an existing 2G cellular sys-tem) or the specific needs (e.g., good data servicesto remote rural areas with little mobility needs).Looking forward, 4G systems are at this point pri-marily just a name. A good working assumption,however, is that 4G systems will be a marriage ofthe best attributes of 3G cellular and packet-basedwireless access systems.

There is still uncertainty about where such stan-dards will be developed. Data standards have beendefined to data mainly in IEEE while cellular stan-dards have come from the ITU. A new group with-in the IEEE, 802.20, has begun to look at highlymobile systems from a data-centric view. At thispoint, it is too early to decide how 802.20, contin-ued improvements to 802.16, and various possible3G follow-on standards will relate to one another.

One other trend worth noting is toward end-user devices that can interact with multiple typesof networks. Clearly, PCs are intelligent enoughdevices that they can easily support multipleradios or flexible radios that will permit them tocommunicate using multiple standards, choosingthe best available network be it 802.11, 802.16, ora wide-area cellular technology such as GPRS.Smaller devices including PDAs and voice hand-sets will also evolve to support multiple networks.For example, a carrier may support a handset thatuses Voice-over IP (VoIP) over 802.11 if it findsitself in a place where such a WLAN is present butswitch to a standard cellular mode otherwise.Considerable technical work is still required todefine the standards needed to make smooth tran-sitions across such disparate networks work well.

System Architecture Issues

Wired versus wireless options

Any wireless access network eventually getsconnected to the wired Internet, so it is importantto understand the role of wireless with regard to

wired parts of a deployment. Wireless systemshave a number of advantages of which the twomost notable are the portability of the end systemsand the ability to deploy without extensive layingof physical wires or fibers. It is important to real-ize, however, that wireless systems pay for theseadvantages: They will never be as high perform-ance as a fiber deployment. Consequently, it is bestto see wireless systems as complementing ratherthan replacing wired ones.

Ideally, fiber would be deployed as deeply into adeployment as affordable and practical. Whenevernew “wires” need to be deployed to carry data,fiber systems are clearly the right choice. Wirelesswould then be used to extend this connectivity toa larger number of locations and to ultimately con-nect end systems to the network. However, sensi-ble deployments will utilize good copper connec-tions that may already be in place to feed wirelesssystems, and wireless access systems may well feedlocations such as offices that may have an internalwired Ethernet system in place. Thus, most real-world deployments will use both wired and wire-less technologies in a cost-effective mix to reachthe most users.

Mesh and multi-hop approachesUp to this point, wireless systems have been dis-

cussed as consisting of base stations or accesspoints (APs) that feed a collection of end systems.But more complex wireless systems make goodsense as well. For example, a simple case wouldconsist of an 802.16 wireless-access system withan antenna on the side of a building that is wired toan 802.11 AP inside the building with which endsystems communicate.

A more complex example is a multi-hop 802.16system. Rather than feeding an 802.16 base stationdirectly with a fiber, one might feed it with a linkfrom another 802.16 base station. Of course, care-ful channel selection is required so that the feedersignal from the first base station does not interferewith the distribution signal sent from the end basestation, but such a system might avoid consider-able construction of fiber links when servingsparsely populated areas. Technical considerations

1.Pages1_End 11/19/03 10:17 AM Page 11


will limit the effectiveness of this approach.Increasing the number of hops will add to the net-work latency seen by the end systems which caneventually adversely impact some applicationssuch as voice.

In designing multi-hop links that will carrylatency sensitive traffic, one needs to define anacceptable latency budget and then compute thelatency added for each relay point. Generally, afew hops can be tolerated without unacceptabledegradation to voice traffic (i.e., latencies worsethan what is experienced by cell users or generalVoIP users today). The specific limits will dependon the specifications of the actual equipment,however. Another limiter to multi-hop deploy-ments is that the traffic from outlying base stationswill fan in eventually to a common link to thehigh-performance fiber backbone, and the totalcapacity of the common links can eventually limittotal system performance.

An even more complex system architecture thatmay be appropriate for some deployments is amesh network. Such designs are still in theresearch stage today but will likely be incorporated

into future standards. In a mesh design all, or atleast very many, end-point nodes also act as relaypoints to other end-point nodes. Essentially trafficto many users is carried through radios at theirneighbors’ homes. This reduces the number ofexplicit base stations that are needed by turningevery end point into a sort of mini-base station. Forcases where the end points to be served are verysparse, this may reduce deployment costs by elim-inating central base stations in favor of addingsmall incremental costs to many user stations. Formore dense deployments, mesh systems canincrease reliability and capacity because there maybe many traffic paths back to the Internet.

An example total architectureWe can summarize much of the discussion

above by considering a figure (above) that illus-trates many of the deployment options. Any realdeployment may only use a subset of the casesillustrated, but all the cases will likely appear insome deployment.

The effective combination of 802.16, 802.11,and fiber and wired infrastructure can support a

AN EXAMPLE TOTAL ARCHITECTURE

1.Pages1_End 11/19/03 10:17 AM Page 12

TECHNOLOGY PRIMER 13

wide variety of cost-effective Internet access rang-ing from traditional office access to remote agricul-tural and community systems supporting ruraldevelopment. It is worth noting that while themobility benefits of wireless access may be viewedas mainly of interest as a luxury for developedeconomies, they may be beneficial even in devel-oping economies as a way to support such activi-ties as development construction, outdoor indus-tries such as farming, drilling, or mining, or evenroving governmental services teams for things likemedical outreach.

The Importance of Standards

Standards are discsussed at length above, and itis quite important to realize how important stan-dards actually are. First, standards create a way forthe best technical ideas from academic and indus-try developers to be combined and amplified. Theprocess of creating the standard subjects the pro-posed ideas to broad review and generally resultsin considerable technical improvements on anyindividual company proposals.

Of great importance to purchasers of the ulti-mate equipment, standards generally result inmuch lower costs much sooner. This is becausethe existence of a standard creates a common mar-ket into which competing companies sell.Furthermore, the size of that common market islarger than any submarket based on proprietaryapproaches would be. In the high-technologyworld, costs are strongly a function of volumes sothe creation of a single high-volume market leadsto much lower costs than a fragmented marketwould. Standards also permit higher degrees ofintegration of equipment, which also contributesto lower costs.

For communications gear, particularly mobileequipment, standards are critical to allowing usersto move around countries and the world while stillhaving their equipment work correctly. While thismay not seem to be of direct importance to the useof communications equipment to connect the citi-zens of developing countries, it will in fact enablethose citizens to more easily become part of theworld economy and benefit from the resulting

development. For example, while mobility may notbe important to providing access to the residentsof a farm, it may be important to the visiting busi-nesspeople who are negotiating for the exports oftheir crops or the introduction of new economicdevelopment projects to the area.

Future Directions

Radio technology and standards are still verymuch in a period of active and successful develop-ment. The performance of the technologies is stilllimited by the amount of cost-effective computingpower that can be applied to the problem. Theproblem of reconstructing a data stream from areceived encoded radio wave is one that is solvedwith complex signal processing mathematics. Asprocessors get faster, more complex algorithmsbecome cost effective and the performance of digi-tal radio systems improves. While there are cer-tainly theoretical limits to such improvements,today there is still quite a way to go to reach them,and therefore we should expect continued tech-nology evolution.

For example, the 802.11 group has begun dis-cussions regarding the next generation of thefamily of standards in a workgroup called802.11n. While it is too early to know what theperformance improvements will be, it is reason-able to expect that the group will target transmis-sion speeds above 100 Mbps and/or improve-ments in the range over which transmissions athigh speed can occur. In addition to continuingto increase the complexity of the PHY encodingschemes, there are also entirely new techniquesbecoming available, such as “smart antennas”that will use processing power to benefit applica-tion performance.

Radios that use the spectrum in different waysare also being considered. One example is a tech-nique called Ultra Wideband, or UWB. In thisapproach, vast bands of spectrum are used atextremely low power to transmit information veryfast but only over very short distances. Theapproach allows multiple uses of the same spec-trum since this UWB signal is so weak that exist-ing users of the spectrum see it as a small amount

1.Pages1_End 11/19/03 10:17 AM Page 13


of noise that can be ignored. In other schemes,smart or agile radios are being considered that canlisten to spectrum bands to see if they are current-ly in use by their official owner, and if not, oppor-tunistically use them to transmit data. Such radioscould allow much more total efficiency in the useof spectrum than our more typical fixed alloca-tions permit. Many of these new radio approach-es, however, will require that regulators redesigntheir spectrum regulations to be more flexible. Itwill be critical for regulators to understand howthey can accommodate such radically newapproaches lest these radios be precluded fromoperating in their countries.

Closer to the application layers, there is alsomuch work ongoing in providing better user expe-

riences. For example, it will become possible tobuild phones that can seamlessly roam from 802.11systems to 802.16 systems to cellular systems. Thiscould permit a single handset to operate with abusiness as an extension but then behave as a cellu-lar phone when its user departs the building.

About the only thing that is certain about thedirections radio technology will take over the com-ing years is that it will become increasingly digital,increasingly “smart,” and continue to change. Thebiggest impediment to the deployment of these newtechnologies may well be the inadvertent barrierscreated by old regulations that preclude deploy-ment. The technology challenges will therefore beshared among the technologists, the regulators, andthose who wish to deploy useful technologies.w

MUCH MORE WIRELESS BROADBAND, FOR MUCH LESS

1.Pages1_End 11/19/03 10:17 AM Page 14

REGULATORY ENVIRONMENT

1.Pages1_End 11/19/03 10:17 AM Page 15

1.Pages1_End 11/19/03 10:17 AM Page 16

REGULATORY ENVIRONMENT 17

A global standard meetslocal policies and politics*

By Axel Leblois

International institutions and national gov-ernments regularly publish announcementsabout the attribution of spectrum portions to

certain applications and service providers. Beyondthe simple sharing of a limited spectrum, however,the challenges facing wireless technology imple-mentation in the developed and developing worldsinvolve many different actors. These include gov-ernments, regulatory agencies, local governments,and in many developing countries, incumbent tele-com monopolies.

Incumbent Telecoms

While developed countries have started orcompleted deregulation of the communicationsmarkets, many emerging economies still run theirtelecom networks through a single, often state-owned organization. These monopolies are thede facto sole providers of technology, transmis-sion, and content.

Over the years, these organizations have invest-ed heavily in wire and cable communications infra-structure and are not yet ready to support wirelessinitiatives, much less to open the way to what maybe perceived as potentially disruptive low-costcompetition. This protective stance is often exac-erbated by financial constraints because wire andcable infrastructure have been financed throughdebt. Quite often, incumbent telecoms also main-tain tight control of national Internet backbone

resources, making it difficult if not impossible fornew wireless Internet service providers to operate.

The pressure exerted by such organizations tolimit the access to new technology and serviceproviders is, in some cases, encouraged by politicalrestrictions on information access. In most cases,the result is a wireless market strewn with proce-dural hurdles that can prevent new actors fromentering the field.

Unlicensed Spectrum

Only 41 percent of developing countries allowunlicensed use of wireless Internet devices and/orspectrum, compared with 96 percent of developedcountries (see Table 1). Most newcomers to thetelecommunications field in emerging economiesface red tape and then a series of fees to obtainlicenses for equipment and/or spectrum use. Forexample, countries charge license fees for hotspotequipment or sites and one-time or annual licensefees for access-point devices.

Most researchers, analysts, and entrepreneurssee the need for a comprehensive regulatoryframework for using limited spectrum. Presently,the unlicensed wireless spectrum was set aroundthe 2.4-Ghz band. In June 2003, theInternational Telecommunications Union madeavailable the 5-Ghz band for license-exempt tech-nology deployment.

The 900-Mhz band, unlicensed in the UnitedStates, is presently used by the international GSMwireless phone standard, used in Western Europeand in many developing countries.

Within those internationally defined zones,Axel Leblois is Co-Founder of the Wireless InternetInstitute and President of World Times, Inc.

*Based on the proceedings of “The Wireless Internet Opportunity for Developing Countries” at UN Headquarters on June 26, 2003

1.Pages1_End 11/19/03 10:17 AM Page 17


national regulations vary widely, with more or lessrestrictions applying to radiation power. What islacking is a comprehensive worldwide policy thatwould align countries and best practices amongindependent wireless Internet service providersaround the world.

On March 20, 2003, in a positive step towardgreater international consistency, the EuropeanCommission issued a recommendation to encour-age member states to provide license-exemptWLAN access to public electronic communicationsnetworks and services in the available 2.4-Ghz and5-Ghz bands. This is not binding on member states,but it is consistent with the (binding) AuthorizationDirective, which requires all member states to allowlicense-exempt access to the spectrum when the riskof harmful interference is negligible. TheAuthorization Directive was scheduled to be imple-mented by Member States on July 25, 2003.

The recommendation is important because itopens the way for using license-exempt spectrumfor commercial purposes, creating large opportu-nities for service providers, and showing how fastchanges have occurred given the amounts recentlypaid by mobile operators for access to the 3Gspectrums. Here, access to the spectrum is free.But it also introduces new issues, which operatorsof radio systems did not have to deal with in the

past, in particular the obligation of noninterferencealready in place in the United States (see p. 20).

Unlicensed Doesn’t Mean Unregulated

Many analysts emphasize that unlicensed doesnot mean unregulated, and all manner of operatorproviding wireless services still needs to maintain ano-interference working plan and a “good neigh-bor” attitude.

Conversely, licensing does not always promotethe development of services. In the UnitedStates, for example, many licenses are attributedto providers that go no farther than the paperthey are printed on! Several countries have earlyon embarked on various options for licensing,such as class licensing and service (as opposed toequipment) licensing. Many of those experi-ments are still under way, and they hold thepromise of opening access to actors in under-served areas through a less cumbersome regulat-ed environment.

Whatever solutions are chosen, enforcing anorderly use of the spectrum and an acceptablequality of the systems may require setting up unbi-ased oversight bodies with legal authority and aclear charter to promote the broadest access possi-ble to users at the lowest possible cost.

ISPs Versus Network Providers

The provision of unlicensed wireless servicesraises further questions because of the two differ-ent kinds of actors in the field: network providersand Internet service providers. Of course, the twomay be the same. When this is the case—that is, theISP and network provider are affiliated—the twotogether will try to curb all competition. In coun-tries where Internet service comes only from theincumbent telecom, competition is effectively cur-tailed, and a fair playing field is very hard to estab-lish. Special interests will do everything in theirpower to hold on to their monopolies. This is par-ticularly problematic when ISPs need to access theincumbent’s network.

Consequently, both Internet-service entrepre-neurs and end users can argue for unlicensing alltogether, letting markets drive both the choice of

% with license-exempt wireless spectrum 96% 41%

% with license-exempt wireless devices 95% 40%

% with license-exempt wireless commerce 65% 20%

Source: UN BCS @ CeBit June 2003

TTaabbllee 11:: Licensing regulations for wirelessInternet in developed and developing countries.

Developed Developingcountries countries

1.Pages1_End 11/19/03 10:17 AM Page 18


technology and the services offered. The rapiddevelopment of advanced wireless Internet tech-nologies is compelling evidence of how robustmarkets quickly create a fair playing field. Bothincumbents and newcomers must then adjust tothe needs of the market, complementing oneanother rather than strictly competing.

Forward Agenda for Regulators

With both the North American and Europeanmarkets taking ever more advantage of wirelessInternet technologies using unlicensed spectrum,it seems ineluctable that wireless Internet solu-tions will gather great momentum and furtherbenefit from economies of scale. Developingcountries have the opportunity to assess whenand how to transition toward similar unlicensed-

Although the issue of voice-over IP is notspecific to the wireless Internet, itdeserves special attention. The develop-

ment of voice-over IP has been extremely rapidin most developing countries, driven by inter-national traffic bypassing traditional circuit-switched accounting and international settle-ments. Consequently, many telecoms have beeninitially concerned about related losses of rev-enues, especially where international calls sub-sidize local traffic. This explains the very strongopposition to voice-over IP expressed at multi-ple occasions by certain developing countries.

At the same time, others have concluded thatvoice-over IP could yield significant savings inhard currency if national telecoms use the serv-ices of international voice-over IP operators.This strategy would eliminate the settlementrate balancing tariffs, saving around 50 percentof the cost of terminating international calls.

Voice-over IP technology has been adoptedby several developing-country telecoms to build

more efficient (lower bandwidth required),lower-cost networks. Examples include Chad,Egypt, Venezuela, South Africa, and Nigeria,with many more in transition. Countries are tak-ing a further step by allowing the licensing of IPtelephony providers. An example is India, whoseregulations once prohibited most commercialforms of voice-over IP to protect its telecommu-nications industry. The net result of these trendsis a likely rapid expansion of voice-over IP indeveloping countries both from an internationaltraffic and domestic network standpoint.

Although the debate on voice-over IP in thecontext of wireless Internet has not yetemerged, it seems that well-designed local reg-ulatory frameworks supporting the deploymentof voice-over IP by local wireless Internet serv-ice providers could potentially accelerate boththe availability of telephony in underservedareas and vastly improve the economics of wire-less Internet by adding voice to data traffic.w

—Axel Leblois

Voice-over IP in potentia

spectrum definitions.To draft a forward agenda, both government

agencies and the private sector should make a thor-ough analysis of market potentials and the technolo-gies that best suit a particular setting. A strong pushfor dissemination of knowledge in the sector seemsindispensable. Internet access in general, and wire-less Internet possibilities in particular, are seldom atthe top of the agendas of government agencies. Acomprehensive catalogue of potential applications,especially in e-education and e-health, is needed toexpand policymakers’ awareness.

The same goes for available technology, whichremains in the somewhat guarded realm of tech-nology wizards. Regulatory bodies need to beenlightened on the options for distance and prox-imity transmission equipments, and on the techni-

1.Pages1_End 11/19/03 10:17 AM Page 19


cal knowledge necessary for implementation andmaintenance. This is particularly true in remoteunderserved areas where the expertise to sustain aworking system may be unavailable. Technicaleducation programs should be delivered byproviders on an ongoing basis, in cooperation withinternational institutions and regulatory bodies.

Last but not least, guidelines are needed to helpassess the economic sustainability of wirelessInternet access systems. Today’s last-mile solutionscan bring access to the most underserved and poorareas. The financing of such networks in itself doesnot constitute a major obstacle to implementation,but the financial aspects of content sale and use are

still pretty much undefined. The legal and economicframework for existing providers and newcomers todevelop those systems should comprised of incen-tives for sustainable and financially sound programs.

A competitive regulatory environment, freeaccess to spectrum within guidelines, alliancesbetween private sector and government bodies,and between incumbent telecommunicationsproviders and newcomers, and an internationalagreement on the basic framework available towireless Internet systems can nurture the develop-ment of the networks and further bridge the digitaldivide. Multi-partner discussions are stronglyencouraged, lest the market take the lead.w

In the United States, regulation of the wirelessInternet now focuses primarily on technolo-gy standards. If manufacturers, practitioners,

and entrepreneurs adhere to technical stan-dards designed to deal with interference, noapprovals are required to operate equipment.You can buy it, take it out of the box, turn it on,and use it. At the same time, because operatorshave no vested rights to continue operation,this constitutes a nonexclusive use of the radiospectrum.

At “The Wireless Internet Opportunity forDeveloping Countries” conference at UNHeadquarters on June 26, 2003, Robert Pepper,Chief of the Office of Plans and Policy at theFCC, touched on the dynamics of unlicensedspectrum, technology regulation, and markets.

“Because of the newer technologies, you cando a lot with unlicensed spectrum. Under ourrules, you may not cause harmful interference.And you must accept interference from some-body else. As a result, the technology companiesare continuing to develop increasingly robusttechnologies that are smart enough to know

when there’s someone else in the neighborhood,in order to avoid interference from someone elsethat you might be receiving.

“You must stop your operation if notified bythe FCC that your device is causing harmfulinterference. From a regulator’s perspective,it’s not only important to establish the base-line rules for the technology, but you alsoneed—and this is important around theworld—to be able to detect people who are vio-lating those technical rules, and have the abil-ity to enforce those rules so you can put themout of business. None of this works unless peo-ple adhere to these rules, and then they canoperate on an unlicensed basis—easy entry,easy exit—using declining-cost technologybased on global standards.

“You also must receive authorization beforemarketing or importing these devices into theUnited States. This all goes to the notion of reg-ulating a device, and not regulating spectrumor use of spectrum. We permit very flexible useof that spectrum as long as your devices adhere

FCC regulates the technology

3

1.Pages1_End 11/19/03 10:17 AM Page 20


The 2003 World Radio-communicationConference finished its nearly month-long ses-sion in Geneva, Switzerland, on July 4, and

took an enormous step forward in promoting unli-censed wireless broadband data services.

At the WRC-03, delegates harmonized preexist-ing, regional allocations of unlicensed spectrum inthe 5-Ghz band, creating a greatly expanded, glob-al allocation of 455-Mhz of spectrum for use byunlicensed wireless networking devices. This deci-sion should significantly broaden the opportuni-ties for people to access information using unli-censed devices, such as Wi-Fi wireless LANs. It isalso likely to enable faster deployment of wirelessdata services in locations where dial-up access isnot practical.

The WRC’s recent vote was the culmination ofa cooperative effort between US government offi-cials and more than 2,500 delegates from countriesaround the world. Representatives from compa-nies such as Microsoft, Intel, Hewlett-Packard,and Cisco worked very closely with US officials inthe months leading up to the WRC to create aframework for spectrum allocation that met theconcerns of incumbent users, reflected global con-siderations, and enabled a global market for unli-censed broadband equipment.

In the case of the radio spectrum and the UnitedStates, Microsoft and its technology-industry part-ners will support the FCC as it codifies the WRCdecisions and allocates the US radio spectrum. Asthis process is mirrored around the world, it willhelp make possible consistent availability of unli-censed radio spectrum that will enable peopleworldwide to access the Internet and to communi-cate locally and globally.

Pierre De Vries, Chief of TechnologyIncubation for Microsoft, predicts the WRC deci-sion will accelerate Wi-Fi as a new broadband plat-form in homes, businesses, and schools in devel-oping as well as in developed nations, resulting in“a rapid deployment of innovative wireless data

and services for business, personal and education-al applications.”

Q. Do you expect this decision to benefitdeveloping countries as well?

A. Certainly. Unlicensed wireless technologycan help developing countries implementInternet networks very quickly, and that has sig-nificant implications for accelerating the growthof information systems in those nations. We arealso seeing rural and suburban applications indeveloped nations taking advantage of Wi-Fi. Wi-Fi is a technology with broad potential aroundthe world, and we think the WRC agreementunderscores that.

Q. Why did the US government and technol-ogy companies come together on this issue?

A. Everybody had a clear sense from the startthat allocating spectrum to unlicensed wirelessconnectivity would have a very positive impact onaccelerating the delivery of broadband networksand data worldwide. This was a goal that both gov-ernment bodies and the private sector felt wasworth achieving, and one that we all workedtogether on.

Q. Is the WRC decision binding?A. Yes, in the sense that any spectrum used con-

sistent with this new primary allocation receivesworldwide interference protection. And equallyimportant, the WRC decision is a strong signal togovernments globally that they should implementthis decision on a national level. We believe thatmost of them will.

Q. In the developing world, isn’t the costof devices and wireless data services still aconsideration?

A. It certainly is, but no one is saying this willresult in millions of Web-enabled devices beingdeployed in the developing world overnight. Butby allocating spectrum in this fashion, the WRChas given governments and the technology indus-try a clear path to growing communication, con-nectivity, and commerce around the globe. Having

Microsoft chief reports on WRC-03

1.Pages1_End 11/19/03 10:17 AM Page 21


everyone moving along a common path is a majorstep forward.

Q. What was the role of Microsoft, H-P, Intel,and Cisco in the WRC decision?

A. We all worked very closely with governmentofficials and WRC attendees. This was clearly a casewhere the industry as a whole benefited from abroad policy decision, and where everyone con-cerned realized that there was a greater benefit thatcould be delivered by working cooperatively on this.

Q. What are the next steps in this process?A. Each government that participated in the

WRC will now go back and work through its regu-latory agencies to codify the decisions made there.This particular decision was one of many deci-sions, so all of the delegates will have lots of

ground to cover with their respective agencies.We’re ready to work with the other industry lead-ers and with the FCC to make the WRC decision areality for the United States.

Q. Will Microsoft be working with govern-ment bodies in countries other than the UnitedStates, as well as with the FCC?

A. Each government must deal with the issueindependently. We believe the WRC decisionspeaks volumes about the need for a worldwidecapability for unlicensed spectrum, and webelieve that all the various international regulato-ry bodies will feel the same way. It truly is a win-win on a worldwide basis for service providers,hardware and software companies, and mostimportantly, for customers.w

to the rules….“It’s hard to predict the future, but we know

it’s already happened. We’re looking at millionsof uses with wireless. The wireless networks in theUnited States are expected to top 15,000 by theend of this year. That’s up by a factor of 15 fromtwo years ago. And it’s not just wireless local areanetwork. We also see various versions and flavorsof Wi-Fi including 802.16 as wireless local loop.

“These technologies are enabling new servic-es in the developing world. In the United

States, we also have our own version of thedeveloping world: very rural areas. We haveseen Wi-Fi technologies being deployed in ourrural areas with stations on water towers andgrain elevators, providing broadband servicesto farmers. There are some very low-densityareas that could never get it before, through lit-tle companies that are breaking even with 400customers. It’s very exciting and it totallychanges the business model. This is why wethink this a very important technology for thedeveloping world globally.”w

3

1.Pages1_End 11/19/03 10:17 AM Page 22

OPPORTUNITIES AND ROADBLOCKS

1.Pages1_End 11/19/03 10:17 AM Page 23

1.Pages1_End 11/19/03 10:17 AM Page 24

OPPORTUNITIES AND ROADBLOCKS 25

By Sandy Pentland

The goal for the Information Society shouldbe broadband connectivity everywhere, ratherthan the village telephone envisioned in many

current programs. While this may seem unrealis-tic, recent advances in technology make it anachievable goal, especially when adopted as partof a progressive, market-driven migration fromgovernment seed services, such as e-governance,to universal broadband connectivity paid for bylocal users.

The key change is that recent advances in wire-less computer networking—particularly the IEEE802.xx standards—have led to huge commercialsuccess and very low pricing for broadband net-works. While these networks are thought of asmainly for offices, we have shown that they canprovide broadband access to even the most remoteareas at very low prices. Today wireless cell-phoneand wireless local-loop, or WLL, service costsabout one third of copper or fiber land-line serv-ice, while packet-based broadband computer net-works cost roughly one ninth of the land-line serv-ice-and have the additional advantage of being farfriendlier to data services, and to lower-grade voiceservice such as voice messaging. These new tech-nologies offer developing countries an opportuni-ty to leapfrog over wireline and wireless-local-looptelephony infrastructure to the forefront of broad-band communications technology.

Wireless data networks based on the IEEE802.11 or so-called Wi-Fi standard are perhaps themost promising wireless technology. The forces

driving the standardization and proliferation of Wi-Fi in the developed world have resulted in featuresthat can stimulate the communications market in thedeveloping world. These features include: its ease ofset-up, use, and maintenance; its relatively highbandwidth; and, most importantly, its relatively lowcost for both users and providers.

Standard Wi-Fi connectivity (the IEEE802.11b wireless standard) provides up to 11Mbps data rates and operates in a band near 2.4Ghz that is generally unlicensed in Europe and theAmericas. Newer versions of Wi-Fi provide 22Mbps in this band, and versions that operate athigher frequencies provide up to 54 Mbps. Testsin rural settings show that a standard Wi-Fi card(as commonly used with laptop PCs) can providegood connectivity up to a half-kilometer radiusgiven line-of-sight. With the addition of antennasand repeaters, it is possible to achieve point-to-point connectivity at distances of up to 20 kilome-ters. Wi-Fi access points (devices commonly usedto provide a Wi-Fi network) currently can be pur-chased for $80 and Wi-Fi cards for under $30.

As one demonstration of the practicality of thisnew technology for rural connectivity, researchersfrom the Indian Institute of Technology at Kanpur,working with Media Lab Asia, have “unwired” a100-sq-km area of the Gangetic Plain in centralIndia. This project provides broadband connectiv-ity to the homes of almost one million people atunder $40 per home. Other experiments haveshown the practicality of the technology in moun-tainous terrain, and in city centers. Indeed, severalcities in the United States have begun to deployfree Internet connectivity using the IEEE 802.11bwireless standard.

Even with advances such as demonstrated in theDigital Gangetic Plain project, the cost of real-time,

Toward a broadband village areanetwork

Professor Alex (Sandy) Pentland is Director of Media Lab Asia, Co-Director of the Digital Nations Consortium, and Toshiba Professor at the MIT Media Laboratory.

1.Pages1_End 11/19/03 10:17 AM Page 25


circuit-switched communications is sufficientlyhigh that it may not be the appropriate starting pointfor rural connectivity. Based on market data forinformation and communication technology (ICT)services in rural India, it has been argued that asyn-chronous service—for example, voice messagingand e-mail—may be a more cost-effective startingpoint for rural-connectivity projects.

MIT Media Lab researchers have developedand demonstrated a combination of wireless tech-nology and asynchronous delivery that reduces thecost of asynchronous Internet service by twoorders of magnitude below land-line expenses.This store-and-forward wireless network for ruralconnectivity is known as DakNet (see p. 79).

The opportunity posed by these technicaladvances and field demonstrations is to redirectexisting universal-service programs away from thecopper-to-village-center model and toward abroadband village-area network serving everyonewithin the village center. Whether beginning withthe inexpensive asynchronous service or the some-what more expensive real-time service, such vil-lage-area communications can be smoothly

upgraded to always-on universal broadband asmarket demand grows.

To achieve this goal, policy makers should beconsidering several policy steps.

The first step is to deregulate the portions of theradio spectrum used by the IEEE 802.xx stan-dards. This spectrum is already deregulated indeveloped countries, which has unleashed anenormous commercial expansion and broughtplummeting prices.

A second step is to permit voice-over Internetpackets, or VoIP. Currently many countries strictlylimit the carriers of real-time voice signal, but nowthat voice can be broken into Internet packets, or IP,these regulations are increasingly archaic andexpensive and limit the growth of digital businesses.

Finally, regulators should consider deregulatingmini Internet service providers, or ISPs. These aretypically Internet-connected computer users, suchas Internet cafés, that allow nearby users to accesstheir Internet connection with an 802.11 wirelesslocal area network. By allowing such connections,the evidence is that there will follow an explosionof connectivity within village and city centers.w

THE DIGITAL GANGETIC PLAIN PROJECT

1.Pages1_End 11/19/03 10:17 AM Page 26


By Iqbal Z. Quadir

Information and communication technolo-gies have some good news for developing coun-tries. These technologies can foster much-need-

ed broad-based economic development, empower-ing people with possibilities that were onceunthinkable. For instance, there are already moremobiles than fixed phones in Africa and, in thenext five years, mobiles are expected to outnumberfixed phones 4 to 1. This rapid expansion of con-nectivity in Africa is surely more attributable to thenew uses and economics of digital technologiesthan to anything else.

I can relate to this from my own experience inBangladesh, where mobiles already outnumberfixed phones 2 to 1 and more and more people areable to coordinate their economic activities moreeffectively. They waste less time connecting withothers and spend more time working together.

I first learned how connectivity can meangreater productivity about a decade ago, when Iwas working in a small investment firm in NewYork City. It was clear that our productivity hadgone up after we adopted a rather rudimentarycomputer network. We began to avoid the cumber-some exchange of floppy disks, and we updatedeach other more frequently as our ideas evolved.

But one day this network broke down, and as Iwas waiting for someone to come fix it, I recalled aday from my childhood, in 1971, when Bangladeshwas at war. My family had left the small town wherewe lived and moved to a relatively more peacefulrural area. One time my mother asked me to getsome medicine for a sibling from someone about

eight miles away. I walked all morning to get there,but when I arrived the medicine man wasn’t there,and I spent all afternoon walking back home.Sitting in New York, I put the two experiencestogether and drew this conclusion: Whether you’rein a modern office or in an underdeveloped village,connectivity is productivity.

To me, this meant that ordinary people could beeconomically empowered through telephones, andthat telephones could become weapons againstpoverty. Connectivity allows people to better coor-dinate their activities and depend on one anothermore effectively. This dependability allows them tospecialize in order to increase their productivity,which can allow them to pay for telephone service.If part of the increased productivity could be chan-neled to pay for the service, a viable commercialproject could be constructed without too muchconcern about existing purchasing power.

Even though telephones are proven productiontools and the increased productivity that resultsfrom using them outweighs the cost of installingthem, virtually all poor countries have a shortage oftelephones, even after a dramatic rise in the num-ber of mobiles. The percentage of households whocan afford a telephone tends to be much largerthan the percentage that actually have one. InBangladesh, in 1993, there were only two tele-phones per one thousand people, and almost allthese phones were confined to a few urban pock-ets. In its rural areas, where 100 million peoplelived, there were virtually no telephones.

This led me to establish GonofoneDevelopment Corp. in New York in 1994 withangel funding. The investors in Gonofone,which means “People’s Phone,” provided theseed funding to help me develop and organizewhat is now GrameenPhone, currently the

For the poor, connectivitymeans economic opportunity

Iqbal Z. Quadir is a Lecturer at Harvard’s KennedySchool of Government and Founder of GrameenPhonein Bangladesh.

1.Pages1_End 11/19/03 10:17 AM Page 27

largest telephone company in Bangladesh interms of number of subscribers.

I realized that a large-scale commercial projectthat could serve all urban areas and 68,000 villagesin Bangladesh would require players not only withlarge resources but the ability to make specificmeaningful contributions in solving the problemson the ground, estab-lishing a real telephonecompany and distribut-ing the services in ruralareas effectively. Afterstudying the conditionsin Bangladesh, I real-ized that a practical anddependable way of dis-tributing the service inrural areas with littlemodern infrastructurewas the key to a realisticand sustainable project.

To solve this practi-cal problem, I felt that amicro-credit programmade sense. Forinstance, small entre-preneurs, backed bymicro-credit fromGrameen Bank, couldretail telephone servic-es to their surroundingcommunities. BecauseGrameen Bank alreadyhad a branch networkthroughout the rural areas in Bangladesh and itsrepayment records were excellent, the Grameenborrower network would be an excellent way topenetrate the rural market. This practical schemenot only convinced Grameen Bank to join in theproject but also attracted Telenor AS, the primarytelephone company in Norway. Eventually, a glob-al consortium emerged that additionally involvedMarubeni Corp. in Japan; Asian DevelopmentBank in the Philippines; CommonwealthDevelopment Corp. in the United Kingdom; and

International Finance Corp. and Gonofone in theUnited States. With the support of these investors,GrameenPhone, established in late 1996, startedbuilding a new cellular network and providingservices to the public soon thereafter.

To date, GrameenPhone has built the largestcellular network in the country with investments

exceeding $300 mil-lion and a subscriberbase of more than amillion. Its rural pro-gram is already avail-able in more than35,000 villages, provid-ing telephone access tomore than 50 millionpeople, while helping tocreate micro-entrepre-neurs in these villages.GrameenPhone is anexample of technologydramatically unleashingnew possibilities and,through specific proj-ects, increasing broad-based empowerment.

Some people thinkthat poor people aretargets for aid, forget-ting that people arethe most importantresource poor countrieshave. Development isnot resource transfer

from rich countries to poor ones but engagement ofcitizens in economically meaningful ways. InGraameenPhone, citizens are not just receiving aservice, they are actively engaged in producing it.

While some people are worried about the pur-chasing power of citizens in poor countries for aservice they may want to provide, citizens’ involve-ment in distributing the service actually lowerscosts and helps bring the service within the pur-chasing power. Moreover, the availability of a serv-ice makes people more productive and raises their


GrameenPhone serves 35,000 villages in ruralBangladesh.

WIRELESS COVERAGE

1.Pages1_End 11/19/03 10:17 AM Page 28


purchasing power.Others think that education is a project separate

from economic activities. But engaging citizens ineconomic activities is an education in itself. One ofthe surprising aspects of the GrameenPhone expe-rience is that the women who are selling the serv-ices generally need no more than two days, andusually one day, to be trained to operate mobilephones. Engaging in an economic activity is aneducation itself, and poor people, being more eco-nomically needy, are better learners when it comesto education that also enhances their incomes.

GrameenPhone proves that the possibilitiesopened up by new technologies and properdesign can make universal access realistic, evenamid poverty and other difficulties. Technologieslike Wi-Fi, or 802.11, among others, are addingyet more possibilities and moving the technolog-ical capabilities to the point that technology nolonger remains an issue. Virtually unlimited pos-sibilities are becoming realistic. The real issuesare the utility of the service, the delivery mecha-nism, and their economics. And here,GrameenPhone may indeed have something toteach, through the emphasis on delivery and theinvolvement of people.

GrameenPhone did not invent any technology.It simply found a new way of deploying a servicerapidly and effectively and in an economicallysound way, while engaging and consequentlyempowering poor people. Entrepreneurs andinnovators need to focus on the delivery mecha-nism and how to engage people, with the likeli-hood that the right way to engage people may alsolead to the right delivery mechanism.

There are, of course, challenges. The up-front costs of connectivity technology can stillbe high for a poor country. For example,GrameenPhone’s initial outside capital was$120 million, and it took five years to assemblethat capital. The bureaucracies in the regulatorybodies also pose a significant challenge. Thesethings require involvement of large players.That is why, for example, shareholders andlenders in GrameenPhone tend to be large. But

actual development happens when smallerentrepreneurs are activated. The smaller in sizeand the larger in number, the better they are fordevelopment. Third, poor infrastructure ecolo-gy poses yet another challenge. In a poor coun-try, other services, whose quality determine theease of introducing a new service, tend to be rel-atively poor.

These challenges are far less daunting in thecontext of Wi-Fi. First, you don’t need big up-frontcosts. One installation can be made with a fewthousand dollars and it can make the next installa-tion easier. The low up-front costs can help theemergence of many small entrepreneurs, who con-tribute to genuine economic development, notonly by spreading the economic benefits but alsoby dispersing power and helping democracy takeroot. The spectrum necessary for Wi-Fi requiringno license further makes it easier for small entre-preneurs to enter the business and pass on the sav-ings to consumers.

Wi-Fi usually requires few up-front regulatorypermissions and can eventually lead to regulationsthat are sensitive to the needs of consumers andproviders. Except in very authoritarian countriesof the developing world, Internet serviceproviders, or ISPs, emerged in an unregulatedenvironment. That is, in developing countrieswhere regulations generally evolve very slowly, theInternet arrived relatively suddenly because devel-oping countries’ governments simply weren’t pre-pared to regulate it. When the regulations didcome, the government needed to be sensitive to theneeds of the entrepreneurs and of the consumerswho were already enjoying the service.

In the case of Wi-Fi, or in general for a tech-nology where many entrepreneurs can emergebecause of the economics of the technology,eventual regulations tend to be healthy.Entrepreneurs participating in the field canargue for their rights, making eventual regula-tions conducive to growth. The relative ease withwhich entrepreneurs can enter the business pre-serves competition and protects consumers, evenwithout regulations.w

1.Pages1_End 11/19/03 10:17 AM Page 29

1.Pages1_End 11/19/03 10:17 AM Page 30

31OPPORTUNITIES ROADBLOCKS

By Theodore H. Schell

The examples of wireless Internet serviceproviders, or WISPs, in the Case StudyCompendium are all works in progress. Each

embodies the hope for economic success as wellas an opportunity to contribute to the social andeconomic development of a country or region.Each shows the potential of wireless technologyas an alternative to the traditional copper andfiber infrastructures that are the dominant meansof providing access to basic voice and advanceddata services.

But the successful exploitation of this new tech-nology is not guaranteed, and multiple roadblockscan present themselves in virtually every context.The question is how can wireless Internet entre-preneurs and interested governmental and non-governmental organizations identify the nontech-nology core issues that may arise in order to devel-op well-grounded strategies to improve success?

The Business Opportunity Defined

All ventures begin by identifying an opportuni-ty through a business plan. For wireless or othertelecommunications infrastructure ventures indeveloping countries and underserved communi-ties, the investment thesis typically runs along thefollowing lines:3Wi-Fi and other wireless technologies are inexpen-sive, within certain performance parameters, relativeto the cost of building virtually any alternative access

plant. Consequently, they enable development ofservice offerings that hitherto would not exhibit therequisite price/performance characteristics.3The operating costs of a wireless-based venturecan also be considerably less expensive if contex-tually appropriate performance levels for reliabili-ty, customer service, and support are established;and if one manages the business to minimizedemands on back-office billing and other supportsystems. This, coupled with the lower cost envi-ronment targeted for service, yields operating costsnot only far below those in developed nations oradvanced metropolitan areas, but also appropriatefor the targeted market.3Sufficient revenues can be generated to yield apositive cash flow. This is typically based on theview that in underserved areas, in both developingand advanced nations, a sufficient latent or unmetdemand for the services exists among those withsufficient disposable income to pay the bill, that thebalance of revenue and costs will assure a profit.3The benefits are assumed to be sufficiently sub-stantial from a political and social perspective,such that the government in general and the regu-lator in particular are assumed friendly and thusthe cooperation of the dominant carrier(s) is alsoassumed; for example, interconnection costs arefavorable, provisioning will be done in a rapid andefficient manner, outages will be expeditiouslyaddressed. In addition, three additional and keyassumptions are typically made:3the target is outside the sphere of interest of

the incumbent carriers, either too small or remoteto command their attention given capital con-straints and more compelling opportunities;3aggressive pricing by the incumbent will not

materialize in response to the challenger as itwould devalue the existing revenue base of the

Wireless Internet venturesin developing countries

Theodore H. Schell is the Founder of CometaNetworks. He is a former Senior Vice President forStrategy and Corporate Development at Sprint andhas been a General Partner at Apax Partners. Heserved in the Carter Administration from 1978 to1981.

1.Pages1_End 11/19/03 10:17 AM Page 31


incumbent, and thus the entrepreneur will be ableto extract a fair margin; and3should anti-competitive behavior come to

pass, the regulator, in support of competition andof the social and economic benefits engendered,will act to assure that predatory or otherwise anti-competitive practices are stopped and the incum-bent penalized severely.

Key Issues in Realizing the Opportunity

If all proves out, everybody wins. Seldom, how-ever, has such happened. Consequently, in today’senvironment, still reeling from the meltdown of theglobal telecommunications industry of the pastseveral years, and skeptical of the overall globaleconomy, let alone that in developing countries ordepressed regions, private investment capital willbe extremely difficult to secure. Assuming project-ed costs stand up under close scrutiny, the keyissues of the prospective investor will be the fol-lowing, because each has occurred a sufficientnumber of times in both developing and devel-oped countries to give pause to even the most risk-tolerant investor. The central questions are:3Will sufficient demand in fact manifest itself ?3Can one really be assured that the dominant ormonopoly incumbent carrier will not retaliate, andthat the regulator will establish and aggressivelyenforce a competitive framework?3Is there a real opportunity for financial liquidity;that is, will the business grow sufficiently large andprofitable to attract a buyer or to enable it to becomea public company so that the investors can realizeand repatriate the value that has been created?

Space allows only a cursory characterization ofeach of these topics—sufficient to sensitize one tothe issue.

DemandSeveral factors underlie skepticism about the

absolute level of demand likely to manifest itself.First, 802.11 technology, the primary tech-

nology capturing the imagination of the entre-preneur, requires that the end user have accessto 802.11 capable devices. Consequently,demand for the service will be held hostage not

simply to the marketing of the service but also tothe diffusion of 802.11-enabled devices into thehands of the potential user, such as 802.11-enabled computers and phones. This cost mustbe accounted for.

In developed economies, 802.11 technology isonly beginning to find its way into the mass mar-ket, though the abundance of computers, PDA’s,and cell phones is such that if one simply assumesthe market rate of customer-premises-equipment,or CPE, replenishment, not to mention newdevices, tens of millions of Wi-Fi-enabled deviceswill diffuse into the market assuring a sufficientaddressable market. As such, the question inadvanced economies is simply one of price andutility. In developing markets, however, it is other-wise. Publicly available kiosks and computingcenters will have to be established, initial comput-er purchases will have to occur as will upgrades ofexisting devices in order that individuals andbusinesses have the technological capability toenjoy the benefits of the low cost telecommunica-tions infrastructure.

Second, while the cost of implementing the serv-ice may well be substantially lower, the environ-ments in which it is being implemented—the targetmarket—is typically smaller in absolute terms, par-ticularly in developing countries, and typically hassubstantially less disposable income available. In allmarkets, the question of the sufficiency of scale toassure profitability will always loom, and prices willneed to be set to reflect this. Consequently, the ques-tion is whether the revenue/cost equation is in factsolved in these markets.

Clearly, the calculus is much more optimisticthan in the past, but the lingering question iswhether or not, or in which places, there is a suffi-cient number of businesses and individuals withthe disposable income to pay the bills, with orwithout public subsidy. It is clear that advancedtelecommunications infrastructure is a sine quanon of economic development; but independent ofdevelopment, there is insufficient resource to sup-port the infrastructure without subsidy. While Wi-Fi improves the calculus, the key question iswhether or not it is sufficient.

1.Pages1_End 11/19/03 10:17 AM Page 32

33OPPORTUNITIES AND ROADBLOCKS

Competitive and regulatory responseThe competitive and regulatory response to the

entrepreneurial venture, even when it is created atthe apparent behest of the regulator or govern-ment, is the most troublesome issue. Personally, Ihave encountered it in four developing countries.In each case, in spite of an encouraging regulatorand a legislated framework for competition, and inspite of an incumbent who was either “uninterest-ed” in the geography being addressed or, alterna-tively, faced a combination of incentives and penal-ties structured to assure “cooperation” with theemerging carrier, the new company suffered majorretaliatory actions that were rationalized orignored by the regulators.

Examples included the following: increasedinvestment and predatory pricing occurred in aregion where the incumbent was chastised by theregulator and in which it pledged not to invest fur-ther; interconnection rates were set untenably highwith the concurrence of the regulator and in oneinstance, interconnection was denied for nearly twoyears; provisioning orders were routinely ignoredleaving substantial areas unable to be served for con-siderable periods of time; billing support, requiredduring transitional periods while data bases werebeing prepared for transfer, was inadequate, and thedata base transfer delayed; and the list goes on. In asimple, Internet-only application, only a subset ofsuch issues will apply, but nonetheless they remainsubstantial concerns.

It is in the very DNA of an incumbent to protecthis franchise and to assure that a new challengerwill not arise, even if the initial activities occur farfrom the major urban and commercial centers.And a start-up, with a lower-cost basis, building abusiness even in a territory of little interest to theincumbent, is tomorrow’s competitor and there-fore an incipient “enemy.” How do they causetrouble? Interconnection with the existing net-work is slow-rolled, circuits are provisioned on ahit or miss basis, outages are repaired, but slowly;quality is not enforced on the leased lines; whole-sale prices are set untenably high; and retail pricesare set to preclude the new service provider from

earning an acceptable margin. Few and far betweenare the entrepreneurial ventures that do not have toturn to the incumbent for essential services; andwhen one does so, it puts the new venture at risk.

This issue has arisen in virtually every advancedeconomy seeking to interject competition into itscommunications market; and it will arise, in allprobability, in every developing country, oftenexacerbated because the clearest way to profitabil-ity for the new venture is to market to the incum-bent’s most lucrative buyers, and he eventuallyturns to this market to hasten revenue growth.

The responsibility to assure the entrepreneur’sopportunity falls to the regulator. A competitiveframework must be put in place that ensures fairand competitive pricing and provisioning of serv-ices from the incumbent to the new serviceprovider, and fair means prices set such that thenew entrant can earn a profit. The regulationsmust be rigorously enforced, for if there is not asteep price to pay, the rules will be violated.

Ensuring competition is not an easy task.Incumbent telecommunications providers indeveloping countries are typically intimately relat-ed to the government if not a part of it. They havepolitical acumen, bureaucratic savvy, and politicaland bureaucratic access. In the name of assuringtheir ability to perform and earn the required for-eign payments and capital returns, they may wellargue against increased competition; in the contextof open markets, in spite of regulations to the con-trary, they may act to assure their own survival, jus-tifying actions after the fact in a politically palatableway. This is not cynicism, it happens. And just asthe entrepreneur must be realistic that such may befaced even in the context of utilizing free, unli-censed spectrum; the government must step up tothis challenge if the benefits of entrepreneurialactivity based upon Wi-Fi technology are to beseized and maximized. It is difficult to count theamount of money that has been lost because thewell-intended promise has not been kept.

Financial liquidityBuilding a good business is, paradoxically, not a

sufficient condition to attract many investors.

1.Pages1_End 11/19/03 10:17 AM Page 33


Often, the question comes down to two additionalissues. Will the business become large enough toattract a buyer; or to enable a public offering,assuming the existence of an appropriate market?Does the political environment enable the repatri-ation of dividends and/or capital gains? Both con-ditions are requisite to attract most sources of pri-vate investment capital.

Conclusion

Market savvy will develop, business plans willbe refined, and many practical, day-to-day operat-ing issues will be overcome. Wireless technologywill assuredly advance, its costs will come downstill further; and, it will be joined by an array ofother technologies that will facilitate implementa-tion of compelling hybrid, cost-effective networks.But two factors emphasized above are outside thecontrol of the entrepreneur and the technologysuppliers, and their appropriate resolution cannotbe taken for granted. They merit repetition.

First, a competitive framework permissive ofexploitation of new technologies and unlicensedspectrum by new, entrepreneurial organizations isa prerequisite; it cannot be compromised. Thegovernmental regulators must establish it andenforce both its letter and its spirit, or entrepre-neurial promise will whither and die.

Second, nontraditional financing must be madeavailable in many markets, for until such time as therobustness of these fledgling enterprises can be con-

vincingly demonstrated, they are likely to be capitalstarved. Will there be exceptions? Yes, and there arecase studies in this volume that speak to them. Butthese are exceptions typically in rapidly advancingeconomies. Enabling the development of these ven-tures in some of the neediest locales will fall to gov-ernments, NGOs, foundations, and the most daringof social investors, a class of “investors” with astrong commitment to the social benefits the exten-sion of a communications infrastructure provides.

The alternative in these developing markets is toforce the incumbent, typically monopoly providers,to initiate wireless-based service in unserved areas,thereby imposing on the incumbent a “social tax”earmarked for telecommunications infrastructuredevelopment and service provisioning. It is posit-ed, however, that this later approach will alwayslead to a suboptimal solution for all.

In industrial countries, the issues of bridging thedigital divide, of spurring economic development inparticular urban and rural areas, or more generallyof assuring access to communications services inunderserved areas, must find a cost effective solu-tion, and wireless technologies are undoubtedlypart of the answer. In developing countries, the verysocial and economic development of the society as awhole depends critically upon such infrastructureadvancement. It falls to us all to work assiduously toassure that the political and regulatory climatereflects these needs, and that institutions step for-ward that are up to the task.w

1.Pages1_End 11/19/03 10:17 AM Page 34

CASE STUDY COMPENDIUM

1.Pages1_End 11/19/03 10:17 AM Page 35

1.Pages1_End 11/19/03 10:17 AM Page 36

37CASE STUDY COMPENDIUM

Left to their own devices: Twelvedeveloping-world case studies

INTRODUCTION

By Noopur Jhunjhunwala and Peter Orne

This Case Study Compendium serves as abest-practices reference tool for the WorldSummit on the Information Society. Over

the next two years, the Wireless Internet Institutewill continue to research and add to an accruingknowledge base focused on growing numbers ofwireless Internet deployments throughout thedeveloping world.

The Case Study Compendium presents twelveexamples, out of dozens received by the WirelessInternet Institute in the run-up to “The WirelessInternet Opportunity for Developing Countries”conference. To prepare the Compendium, theprinciple author interviewed project stakeholders,including field practitioners and funders, to pres-ent a coherent picture. Each case study was inves-tigated and analyzed using these parameters:3mission and goals3accomplishments3technology used3funding sources3collaborations3regulatory environment3expectations vs. outcomes3sustainability3scalability and replicabilityWhether motivated primarily by development

objectives, for profit, or for the challenges theseprojects present, the practitioners shared the goal

of bridging the digital divide using broadbandwireless technology. Each found a unique solutionfor overcoming roadblocks, bringing more hopeand opportunities to less developed regions.

The case studies group into five categoriesrevealing core aspects of wireless broadbanddeployments in developing countries:3Remote Regions provide the harshest environ-mental test of the technology and of the implemen-tation models.3Wireless Internet Service Providers are prima-rily driven by profit but can also contribute toregional development.3Shared Access provides connectivity to regionsthat cannot afford individual access.3Adaptive Technologies rethink off-the-shelfsolutions to overcome local limitations.3Rebuilding Nations can draw from the modelsabove to rebuild a communications infrastructureand jumpstart social and economic development.

In each case study, an analysis box (Things toRemember) highlights key aspects, and aschematic (Network Set-Up) shows how the net-work has been or will be deployed. The com-pendium as a whole raises five sets of issues sur-rounding deployment of wireless Internet in devel-oping countries.

Why wireless broadband?

Because relatively low investments in equip-ment can be used to connect a very large area, wire-less broadband can be the most affordable last-mile solution available. For two reasons, it is oftenthe only practical solution. Environmental limita-

Noopur Jhunjhunwala is Case Study Manager for theWireless Internet Institute. Peter Orne is Editor of TheWorldPaper, World Times, Inc.

1.Pages1_End 11/19/03 10:17 AM Page 37


INTRODUCTION

tions, such as moving glaciers or deep rainforestinteriors, curb the ability to physically lay fiber onthe ground. And in some poor countries, wherecopper or wire are valuable commodities and vul-nerable to theft, wireless provides a tamper-proofsolution. Moreover, the networks can be deployedquickly, and an administrator can track a problemat any point, enabling fast repair. It’s easier toreplace an antenna or correct a line-of-sight than tolay new copper or fiber.

Network considerations

The technology in a network is only part of thepicture. Critical is how it is set up, what devices areconnected to it, and how it is powered. The devel-oped world uses wireless mostly for indoor localarea networks, but developing countries are usingit largely for outdoor wide area networks.Consequently, in developing countries, practition-ers need to improvise with off-the-shelf equipmentto widen the coverage area, often with amplifiers.Today, wireless broadband coverage can extend asfar as 50 kilometers.

Because equipment used to connect to the net-work is still expensive, practitioners often mustdepend on low-end configurations or donateddevices, making do until efforts to bring costsbelow $100 are successful. Another considerationis that extreme temperatures and harsh weatherconditions require improvisation.

Availability of power is often taken for granted,

and the cost of having a power-supply back-up canbe underestimated. Consequently, systems may godown for long periods, and a project can run overbudget. It’s important to look at all possible alter-natives, such as solar power.

Search for sustainability

Whether the project is for profit or not-for-profit, achieving long-term sustainability can crit-ically depend on local partnerships, hands-ontraining of local technicians, and buy-in from endusers. In a development context, building anunderstanding of the Internet and awareness ofwireless broadband advantages (over existinginfrastructures) can generate buy-in from the com-munity. Determine up front the range of needs ofend users, involving them from the beginning tocreate a sense of local ownership.

Shared access through community centersinvites a wide range of additional users while lever-aging existing infrastructure—and potentiallystaff—at local post offices and universities. It canalso attract nongovernmental organizations andtheir supportive resource bases.

To move from dependence to self-reliance, thesearch for sustainability should begin early, withcareful thought to potentially generating revenuesfrom the project through small fees for online serv-ices and training, or by developing offline projects,such as ecotourism, to subsidize connectivity costs.

In for-profit deployments, it is important, with

1.Pages1_End 11/19/03 10:17 AM Page 38

CASE STUDY COMPENDIUM 39

INTRODUCTION

some exceptions, not to over-invest and scale uptoo fast. Let the business grow gradually as newdemand is identified, often by word of mouth.

Applications are everything

When people log on to the Internet, they want tocommunicate, and therefore e-mail and voice andvideo mail are easily the most popular uses of wire-less broadband networks in the developing world.Local people want to get in touch with relatives andfriends who may have migrated far distances foremployment or education opportunities. The Webcan make this easier in a way that national telecomsnever could. Internet telephony, or voice-over IP, isalso popular in countries where it is permitted.

The Internet’s biggest advantage over othercommunication technologies is that it can providean array of solutions and products, including thethree “e’s”: e-governance, e-health, e-education (afourth is e-commerce). Telemonitoring devicesenable doctors to diagnose patients living thou-sands of miles away. Distance education programsteach reading and writing. E-governance initiativesenable villagers to update records and file com-plaints from their village, saving a day’s pay. Thelist goes on and on.

Like technology, however, applications also facechallenges. Local scripts require special software.Illiteracy requires voice solutions. Online servicesrequire offline resources—teachers for e-educa-tion, doctors and nurses for e-health, payment and

postal systems for e-commerce. The virtualdemand chain may be established, but it is mean-ingless if the offline supply chain is missing.

The regulatory master

Incumbent telecoms in the developing worldguard their terrain carefully. In rural areas, whichare marginal markets for incumbents, entrepre-neurs may face few initial barriers to implementa-tion. In all cases, it is important to know the rules,seek required permits at both local and nationallevels, and try to bring the incumbent in on theplan, which could enable future scalability.

In cities, the entrepreneur is likely to face theincumbent and regulatory authority head on.Different responses to these entities, however,present varying risks. One is to directly confronttelecoms and authorities wholesale, lobbyingagainst them domestically and internationally,which can lead to confiscations, project termina-tion, and even criminal charges. This routerequires courage, patience, money, and time.

Another option may simply be to bypass the reg-ulatory environment altogether if the authority is nottoo aggressive. Still another is to engage in thebureaucratic process, get required permissions, andeven try to form an alliance with the incumbent.Authorities may be more convinced by projects witha social-development outcome. A final option is todesist until the regulatory environment changes. Butin waiting there is little to be gained.w

1.Pages1_End 11/19/03 10:17 AM Page 39

1.Pages1_End 11/19/03 10:17 AM Page 40

REMOTE REGIONS

1.Pages1_End 11/19/03 10:17 AM Page 41

1.Pages1_End 11/19/03 10:17 AM Page 42


REMOTE REGIONS

Bringing Internet access to the base campof Mount Everest may well rank among theseverest tests of wireless-broadband tech-

nology in the world.Among the well-noted challenges of the

Nepalese Himalayas—high altitudes, extremelyharsh weather, no roads—is the lack of functionaltelecommunications infrastructure. The roughmountainous terrain of the Khumbu Region makesdeployment of cables extremely difficult and cost-ly. Consequently, wireless broadband is a preferredsolution for providing Web and voice-over IP serv-ice to both inhabitants and visitors.

But it’s not easy achieving and sustaining suchservice at the base camp to Mt. Everest, which isaccessed with a four- to six-day walk from the vil-lage of Lukla. Because the altitude is around 5,600meters, one cannot simply land by helicopter.

There is no power, no buildings, and no infra-structure of any kind in this region. The base campis located on a glacier that moves about 10 feeteach day, prohibiting satellite installation. On topof this, temperatures fall to -20°C at night and canreach 15°C during the day, taxing batteries andelectronic components.

The Himalayas, however, are about rising to thechallenge. Thanks to a partnership between theSagarmatha Pollution Control Committee andWorldLink Communications of Nepal, on April14, 2003 the Linking Everest project did just that.With a goal of operating wireless Internet accessfor one tourist season, a six-man “Virtual Yetiteam” led by Tsering Gyaltsen Sherpa wirelesslylinked base camp to a server in Kalapathar using aCisco Aironet 350, and from there to the Internetvia satellite.

Rising to the challenge33Linking Everest, Khumbu Region, Nepal

THINGS TO REMEMBER3Can you imagine digging through a moving glacier to lay fiber? Because of cost, rugged terrain, and deploy-ment logistics, in remote regions wireless connectivity is often the only telecommunications option.3In most remote areas, access to basic supplies can delay the connectivity process. For example, in theKhumbu Region, obtaining a single tool can take a four- to six-day walk, cutting into project deployment timeand future revenue generation.3Regular maintenance is required for smooth and continuous operation of a wireless facility. In Nepal, thetechnician got altitude sickness, causing the base camp to lose connectivity. Qualified personnel amust beavailable in a sustainable way.

1.Pages1_End 11/19/03 10:17 AM Page 43


REMOTE REGIONS

Among other innovations, the team needed todesign special Styrofoam cases for outdoor equip-ment to conserve heat. Bucking conventional wis-dom and popular belief, it was able to operate aVSAT terminal on the glacier.

“This isn’t commercially motivated,” saysDileep Agrawal, founder of WorldLink ISP, whichprovides capital, equipment, and maintenance forthe project. “We got involved because we found itvery fascinating. We want to help this community,and we want to prove to the world that we can con-nect Mt. Everest.”

There are two seasons on Everest, and duringthe second, September 15 to the end of November,the technician manning the base-camp station gotaltitude sickness and needed to come down, lead-ing to a temporary closure and a revenue shortfall.

“It’s all up in the air now,” Agrawal reports. “Weneed a full-time person as a technician.” Findingthat person is a major barrier to continued successand could kill the hope of sustaining Linking

Everest over the long term. Currently, WorldLinkis training a local Sherpa more suited to the base-camp conditions, but low literacy and awarenessamong the local population are obstacles.

The project has several long-term goals, one ofwhich is to link Namche, the first major settlementon the four-day trek from Lukla to the base camp.In the process, small villages, monasteries, andtourist spots will also be connected and thenlinked to the Internet using satellite. Expected usesare e-mail and Web access, voice-over IP, distanceeducation, and telemedicine.

At a very basic level, there is a fundamentalneed among the population to communicate.Many villagers in the region have family inKathmandu, but government phones aren’t easilyaccessible, and there’s never any guarantee theywill work. Internet access could enable lessreliance not only on the phone system but on thefeeble postal service.

Cyber cafés already operate in Namche.

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 44


REMOTE REGIONS

Currently, an attendant types and sends e-mail forthe locals, who seem to have little problem com-municating with their relatives. There is anassumption that relatives have access to e-mail. Ingeneral, Sherpas who have moved from the regionare better off, technically aware and educated.

Proliferating Internet use in the region itself,already traveled with well-heeled tourists, is expect-ed to gradually spur further economic and socialdevelopment locally. The idea is to charge tourists$1 per minute for Internet use and VoIP and useprofits to support local use of the infrastructure dur-ing months when tourism slows. During the firsttourist season, March 15 to June 1, the projectraised $17,532, generating a $10,032 profit after

deducting $6,250 for bandwidth and $1,250 forwages and miscellaneous expenses. This revenuewas sufficient to pay the operating costs untilSeptember 15, when the next tourist season begins.

Dave Hughes, another project supporter, isworking to connect the Sir Edmond HillarySchool in Namche using wireless broadband.There’s also an effort to provide English classesover the Web. One Sherpa who moved toColorado is trying to give instructions in English.

For now, the Linking Everest team needs onceagain to reestablish connectivity at base camp, con-nect the school, and expand its network. But theharsh challenges of the Khumbu Region will con-tinue to test both technology and resolve.w

Tsering Gyaltsen Sherpa logs on wirelessly at Icefall,Everest base camp.

Linking Everest charges $1 per minute at the base campcyber café.

A map of the Khumbu Glacier basin shows theKalapathar/Gorek Shep area, base camp, and Everestsummit.

The Virtual Yeti team wirelessly linked Everest base campat an altitude of 5,600 m to a server in Kalapathar at5,450 m.

1.Pages1_End 11/19/03 10:17 AM Page 45

1.Pages1_End 11/19/03 10:17 AM Page 46


REMOTE REGIONS

Individuals and institutions concernedabout the fate of the Amazon and its Indianpopulations have in recent years begun to

lend a helping hand. In the 172,000-hectareXixuaú-Xipariná Ecological Reserve 40 hoursupriver from Manaus, the welfare of the CabocloIndians is taken very seriously by the AmazonAssociation, an NGO based in Brazil and Italythat views the Caboclo as key to the preservationof the rainforest.

The Amazon Association has teamed up withthe nonprofit Solar Electric Light Fund inWashington, DC, to put broadband Internet at theservice of the Indians, who lack access to basichealthcare, education, and economic opportuni-ties. E-mail, telemedicine, online education andeven e-commerce promise to radically improvetheir lives.

Solar panels now power a satellite Internetlink in the reserve, which enables the AmazonAssociation to maintain permanent contact withIndians who have learned to use e-mail, and whocan report on problems at the reserve, participatein decision-making, and request supplies andmedicines.

“It is important to upgrade the ICT capabilitiesof the people, and build their knowledge base,”says Chris Clark, founding member and currentpresident of the Association.

“The mere ability to communicate on a regularbasis with the outside world has provided atremendous psychological lift to the communitythere,” says Robert Freling of the Solar ElectricLight Fund.

Solar energy provides the electricity, a two-wayVSAT from OnSat Network Communications

Rainforest IP33Xixuaú-Xipariná Ecological Reserve, Jauaperi River, Brazil

THINGS TO REMEMBER3Alternate, cheap, and reliable sources of energy often go hand in hand with wireless-network deployment.Because of the total lack of electricity in the rainforest, solar power has been critical to this project.3When a project is funded entirely by grants, it is key to work toward sustainability from the very beginning.After the system is up and running, ecotourism and e-commerce may help pay for equipment, maintenance,and expansion.3The telemedicine component is a vital-signs monitor loaned to the reserve and administered remotely byUS doctors. Consequently, equipment and labor costs could impede scaling this component to other rain-forest communities.

1.Pages1_End 11/19/03 10:17 AM Page 47


REMOTE REGIONS

enables high-speed connectivity, and a wirelessLAN makes access flexible. The project hasapplied for a spectrum license from Anatel, Brazil’snational telecom.

Of the $75,000 project budget provided by theErnest Kleinwort Charitable Trust, $50,000 wentto hardware and $25,000 to planning, training,and management. The economics of the project,however, remain a serious barrier to growth.

With an eye toward achieving sustainability, agovernance board has been set up to collect all rev-

enues from the project in a common account andwill decide how best to allocate the resources. Theproject is expected to sustain itself within the nextfour to five years. It is also setting up a model farmin the reserve so that the community membersbecome more self-sufficient in food production.And there is currently discussion about using solarenergy to power a Brazil-nut processing facility inthe reserve, which will help generate income forthe community.

The project does have a modest e-commerce

Grants and good will power the project.

NETWORK SET-UP

Solar panels power the wireless Internet.

1.Pages1_End 11/19/03 10:17 AM Page 48


REMOTE REGIONS

“This project was first conceived in November2000 when Lady Madeleine Kleinwort intro-duced me to Christopher Clark of the AmazonAssociation. Working together, we developed aproject proposal and submitted it to the ErnestKleinwort Charitable Trust which awardedfunds in June 2001.

“After severalmonths of planningand preparation,equipment wasordered and airshipped to Manausin late January2002. I traveled toManaus with a rep-resentative ofOnSat NetworkCommunications, asolar technicianresponsible forinstallation andtraining, and adocumentary film-maker. As it turnedout, the twoGateway computerswe had ordered aspart of the shipmenthad been stolen, butinstead of releasingthe rest of the cargo, customs officials at theairport in Manaus kept everything locked up.It took an entire week to liberate our ship-ment, and the whole project almost came to ascreeching halt!

“Once the equipment was in our possession,everything was loaded onto a large riverboat

owned and operated by the Amazon Associationfor the 40-hour trip up the Amazon. Becausethe water level on the Jauaperí River was lowat the time, the riverboat was unable to make itall the way to the reserve, so we had to transferall the equipment onto outboard canoes for the

final several miles.“Altogether, it took

four days to installthe solar-powerequipment, lights,refrigerators, fans,computers, and satel-lite system in theXixuaú. Using aGlobalstar satphone,Kuen Damiano fromOnSat was able tocommunicate withthe satellite operatorin Mexico to lock inthe signal providedby SatMex V. Shortlythereafter, the wire-less LAN was inplace, and we wereable to send andreceive e-mails fromthe middle of theAmazon rainforest.

“Given availabili-ty of additional funding for equipment andtraining, there is no reason that this projectcould not be replicated in many other parts ofthe Brazilian Amazon; indeed throughout theentire developing world, at least on a noncom-mercial basis.”w

—Robert Freling

In Manaus, minus two computers

Equipment delivered by river.

Delivery of data by satellite.

1.Pages1_End 11/19/03 10:17 AM Page 49


REMOTE REGIONS

Remote diagnosis in the Amazon

Thanks to a collaboration with theUniversity of East Carolina TelemedicineCenter and the University of South

Alabama, the Amazon Association can nowuse a computer and a vital-signs monitor toremotely take the blood pressure and performan EKG on the Caboclo Indians, allowingdoctors in the US to provide medical consult-ing and diagnosis.

In one telemedicine experiment just gettingunder way, a nurse hooks up patients to a“Vital-Link” device which is used to measurethe blood pulse, blood pressure, blood oxygenlevel, body temperature, and EKG of everyonein the community. A doctor at the University ofEast Carolina Telemedicine Center has controlover the machine and can even inflate an arm-band to take blood pressure. The informationis transmitted over the Internet to a doctor, whothen videoconferences with the nurse to delivera diagnosis.w

component: The women of the local villages havebegun taking craftwork orders from foreign visitorsguided to the region by the Amazon Association.They recently received a $300 Internet order froman Italian visitor.

The Amazon Association has built three newschools; one is even connected to the Internet.Other projects goals include distance learning,

eco-tourism, biodiversity research and mapping.“The plan now,” Freeling says, “is to extend con-nectivity to other communities along the RioJauaperí by means of Wi-Fi technology.”

“The Association will never abandon thisgroup,” says Clark, who has committed to mak-ing the reserve and this link to the Caboclo a life-long project.w

1.Pages1_End 11/19/03 10:17 AM Page 50

WIRELESS INTERNET SERVICE PROVIDERS

1.Pages1_End 11/19/03 10:17 AM Page 51

1.Pages1_End 11/19/03 10:17 AM Page 52


WISPS

Malawi’s copper-based telephone net-work is often busy and frequently breaksdown, and fiber-line use is limited only to

areas where economic activity already exists. Thecountry’s two cellular operators are far too expen-sive for general data services or even rural voiceservices, yet establishing an Internet link inMalawi—the equivalent of a T1 line—can run to$10,000 a month.

These are tough obstacles facing any entrepre-neur wishing to deploy a profitable data communi-cations network. But Malawi is also a LeastDeveloped Country, seeming to offer little hope forprofitability.

Thanks to wireless infrastructure, however,Africa-Online in Blantyre has managed to providebroadband Internet service in Malawi for morethan three years, and it’s still going strong. Fromthe outset, the company committed to achieving areturn on investment, targeting areas where service

could be economically sustainable, such as theeconomic centers of Blantyre and Lilongwe.

“Conditions are extremely harsh for this type ofequipment, and only commercial motivation keepsour network operational and well maintained,”says CEO Paul Shaw.

Bandwidth goes from a satellite to customersthrough Africa-Online’s wireless network, requir-ing careful economic planning in the back office.“We have to pay about 100 times more for band-width than an ISP in the United States or Europe,”Shaw says. “This alone is the single-most impor-tant factor in modeling the business and has a mas-sive downstream effect on our market.”

Shaw estimates Africa-Online’s current value at$500,000, as a going concern, after an initial share-holder investment of $100,000. No dividend hasbeen paid, and all earnings have been ploughedback into the company. Africa-Online is the onlyone of four to five ISPs in Malawi providing full-

Slow incubation success33Africa-Online, Blantyre, Malawi

THINGS TO REMEMBER3It is important not to over-invest and outpace demand by expanding too rapidly. Africa-Online emphasizesthe virtues of starting small and growing organically for the long term.3Installing technology is not enough. It is important to reserve resources for training customers how touse it.3A wireless connection cannot be stolen. Where copper-wire infrastructure is routinely resold as raw materi-al, wireless proves its advantage in Malawi as it does in rugged or mountainous terrain.

1.Pages1_End 11/19/03 10:17 AM Page 53


time wireless access to 200 of its 2,000 subscribers(most of whom are dial-up).

Africa-Online says the secret to success is startingsmall and growing organically, mostly by word ofmouth, making minimum investments for minimumreturns. “Major investment in high-tech equipmentin developing countries is a sure-fire road to disas-ter,” Shaw says. “It is not practical to create an infra-structure and then wait for clients to line up to useit. Allow years for the learning curve to kick in.”

Africa-Online found it had to provide networkinginstallation and services, routing and switching serv-ices, and training on mail exchange and firewalls. “Itis not enough to facilitate the installation of a specif-ic technology; you have to support all aspects anduses of that technology and be able to do it on acommercially sustainable basis,” Shaw says.

He says it is important to take great care inselecting equipment, and not to assume the indus-try standard is the best in every circumstance.Africa-Online uses Alvarion BreezeNet equip-ment. “Alvarion provided initial training but wehad to improvise,” Shaw says. If equipment fails,

Africa-Online will replace it for no cost throughthe Israeli manufacturer.

Incubating economic development

Seventy percent of Africa-Online clients arecommercial, and Shaw says he has seen a positiveimpact from broadband services on Malawi’s cor-porate and banking sectors. High-speed Internetaccess has brought confidence to externalinvestors and helped rural agricultural clients, forexample, get access to commodity pricing to max-imize their returns. Because of the poor state of thetelephone network, e-mail is now the primarymedium for doing business with both national andinternational clients.

“Clients first want e-mail connectivity, thenwant to be able to research their businesses, andlastly the businesses want to have a presence on theInternet so they can get some sort of economicreturn from this,” Shaw says.

Some businesses with Web sites hosted by Africa-Online have begun to increase their exports becauseof their ability to take online orders. At least two agri-

WISPS

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 54


WISPS

Africa-Online enjoys new offices in Blantyre. The company originated in 2000 in this A-frame.

cultural organizations are exporting tea and coffee byselling and shipping directly to customers in Kenya,the UK, and South Africa. While the Internet mostlybenefits large farmers who can afford to finance ship-ping and absorb payment delays, the hope is thatsmall-scale farmers will form cooperatives and bene-fit from their own Web presence.

Africa-Online offers residential users discountrates at night (6:00 pm to 6:00 am) at a fraction ofthe normal cost. If regulators approve, the compa-ny will gradually move into voice-over IP. The reg-ulatory position on VoIP is less clear but has notbeen ruled out. The ISM bands are available foruse after payment of a registration fee.

While it has not brought Internet access to theurban or rural poor, Africa-Online has given donors

and NGOs access to better and more efficient tech-nologies, assisting them in the delivery of develop-ment solutions. “We can see that the introduction offull-time access to the Internet has led to greaterknowledge being available in the community at largeand increasing interest in our services,” Shaw says.

An agricultural concern has queried the compa-ny about providing connectivity to a farming areain the south of Malawi, given that the state telecomonly offers a 19.2 Kbps analog connection. Africa-Online will consider expanding to rural areaswhere willing entrepreneurs could run cyber cafes.In addition, they have proposed connectingMalawi’s postal service, replacing telex with e-mail, and replacing a “letter writer” with an “emailtypist” in every village.w

It has connected a local backpacker lodge. Africa-Online’s installation for an Internet café.

1.Pages1_End 11/19/03 10:17 AM Page 55


WISPS

Providing broadband in a Least DevelopedCountry with fewer numbers of subscribersrequires a careful study, says Africa-Online

CEO Paul Shaw. It’s not just about providing anaffordable last-mile solution.

“For a symmetrical T1 leased-line connection tothe Internet in the United States, prices vary from$750 in, say, California to $1,000 in rural areas.But that same bandwidth in Africa can cost from$24,000 to $55,000. If I am lucky enough to beliving in a country that allows me to use my ownsatellite equipment, then my costs are at the lowerend of that scale.

“But there are many countries in Africa wherededicated bandwidth is not even available andshared bandwidth must be used with a resultingdecrease in quality—though not much of a decreasein price. The cost of leased lines to customers, or topeering points, is also many times more expensive.

“On the income side, the picture is just as bleak.My company charges $35 per month for an unlim-ited-access dial-up connection, for example.Considering that my input costs are 30 timesgreater than an ISP in the United States, that doesnot compare badly.

“Market sizes are not large. There are only1,600,000 dial-up subscribers in all of Africa, andhardly any cable or DSL subscribers, largelybecause of the extortionately high telecom costs.There’s very low average income—$1,600 percapita per year—in sub-Saharan Africa, and verylow teledensity and low literacy rates.

“On the teledensity issue, Kenya, one of the mostInternet-active countries in Africa, only had atotal of some 450,000 lines installed. Ghana has250,000, a lot less than a large town or small cityin the United States.

Fighting for users“In Africa, a big ISP has a few hundred thou-

sand dial-up subscribers, or a few hundred leased

lines. There are only a handful of these, maybethree. For the majority of African ISPs, 1,000 sub-scribers is a lot. In Mbps, here are the numbers forinternational bandwidth (2002) by country:

Egypt 535South Africa * 399Morocco 136Algeria 83Tunisia 75Senegal* 60Kenya* 28Gabon* 16Nigeria* 15Botswana* 14

*Sub-Saharan Africa

“The remaining countries have less than 13Mbps each, with Equatorial Guinea at the bottom ofthe list with 64 kbps for half a million people. Anaverage ISP in the United States probably has morebandwidth than many of the countries on this list.

“The high cost, relative to income, of being anInternet subscriber in Africa means that many ISPsmust share a very small number of subscribers. Thehigh cost of operating as an ISP in Africa meansthat margins are small to nonexistent, and that it’simpossible to expand the way you’d like.

“Finally, there is the “Half Way Proposal” (seehttp://www.afrispa.org/initiatives.htm), whichcompares Internet provision with voice provision.When an international call is placed from theUnited States to Malawi and another from Malawito the United States, the two telecoms involved havea monthly settlement arrangement where the cost ofterminating the calls is taken into account. Becausefar more calls terminate in Malawi than originatethere, this monthly settlement is a major source ofrevenue for the telecom and of foreign exchange forMalawi. When an e-mail is sent from the UnitedStates to Malawi or from Malawi to the UnitedStates, the entire cost of transit, in either direction,is born by the African ISP.” w

Paul Shaw on beating the backhaul cost

1.Pages1_End 11/19/03 10:17 AM Page 56


WISPS

Making it at the margins33Baja Wireless, Ensenada, Mexico

The smaller cities of Mexico still lackreliable access to basic telecommunica-tions, let alone fast Internet access. Only

44 percent of Mexico’s population has a landline,and it can take Telmex, the national telecom, up tonine months to install basic fixed-line service incertain areas.

But with Internet activity booming in Mexico’slarge cities and in the nearby United States, andwith computer use growing everywhere, smallerworking cities—marginal markets with a good mixof residential, industrial, and commercial activity—are understandably thirsty for broadband access.

In a prototype operation in Ensenada, about 65miles south of San Diego, Baja Wireless is deliveringconnectivity via a 7,500-acre hot zone. With an ulti-mate goal of connecting about 30 cities in the next30 months, this wireless ISP is targeting small citieswith 50,000 to 500,000 residents who may have nowired telephone service because of backlog, whowant or need improved Internet access speed, orwho simply want the convenience of wireless

Internet access. Baja Wireless offers different pack-ages for clients based on their bandwidth usage.

The business model met with skepticism atfirst. “Baja Wireless repeatedly heard howls orblank stares of disbelief at the thought that any-body might be interested in investing in a telecom-munications project in Mexico,” says cofounderSharron Tate, an American. “Relatively few peoplecould see beyond the risk of investing in a foreign,if neighboring, country.” Today, the companyclaims $300,000 in private investment.

After reaching its first 100 customers, BajaWireless determined that its Alvarion hardwarewas cost prohibitive for the mass market inMexico, so it switched to less expensive D-Linkunits and met with great success over large dis-tances and uneven terrain. The switch, however,cost the company a four-month delay, drainingcapital and imparting a valuable lesson. “Identifysources for more capital than you think you willever need,” Tate says. “The unexpected alwaysappears, so be prepared for it.”

THINGS TO REMEMBER3One way to deal with a strong incumbent monopoly is to deploy in marginal markets where the incumbent’srevenue stream is not affected. This might even make collaborating with the monopoly for potential supportand future scaling possible.3Choose equipment carefully. Baja Wireless’s switch from Alvarion to D-Link resulted in better performance ata lower cost, but the transition drained capital and slowed expansion.

1.Pages1_End 11/19/03 10:17 AM Page 57


WISPS

In addition to e-mail communication, BajaWireless anticipates a range of uses across its net-work. A doctor in Ensenada can now share X-rayswith a colleague at Stanford Medical Center, whichwas previously very difficult at dial-up speed. Cash-based businesses can install wireless Webcams toremotely watch cash drawers, and manufacturerscan monitor manufacturing activities.

In the incumbent’s back yard

Baja Wireless describes the regulatory environ-ment in Mexico as “somewhat stringent.” Yet,Telmex, one of the world’s most powerful telecommonopolies, has encouraged its activities.“They’ve been extremely supportive and apprecia-tive and are doing everything to help this initia-tive,” says cofounder Tracy Blackett.

While Baja Wireless is a licensed ISP, it is pro-

hibited in Mexico from promoting or providingvoice-over IP. Telmex charges $.87 per minute toplace a call to the United States and requiresother long-distance operators to charge similarlyhigh rates.

“This is a limitation we have,” Tate says. “ButBaja Wireless got in under the wire just two weeksbefore licensing requirements tightened in Mexico.”Anticipating the company’s growth to other townsand regions, it has contracted with Telmex to use awireless local loop. In this way, the goliath monop-oly becomes part of their overall solution.

Costly bandwidth arrives at Baja Wireless’s officethrough Telmex, which receives it from Sprint at theUS-Mexico border. “In the United States, we wouldpay $300 to $500 a month,” Blackett says. “Here wepay $2,500 a month for a connection which is 25percent greater bandwidth than a T1.”w

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 58


WISPS

It takes a hemisphere33E-link Americas

Wireless Internet service providers atwork in regions with insufficient back-bone connectivity often struggle with

slow growth. Because broadband connectivitycosts are high and revenues are slow tocome, it is important not to scale uptoo quickly and run the risk of over-investment.

But a public-private partnershipcalled E-Link Americas is convinced that abroad-scale deployment of wireless broad-band is a better model, thanks to new satellitetechnology—so-called broadband in the sky—thatmay make it possible to leverage economies of scaleand dramatically lower Internet access fees acrossvast regions of the developing world.

“There’s a specific problem that needs to besolved in the Americas, which is low-cost access inrural areas,” says Randy Zadra, Managing Director

of the Institute for Connectivity of the Americas, orICA, in Ottawa—one of the funders of the projectalong with the World Bank, the Organization ofAmerican States, and several satellite companies.“Picture doctors going to small clinics where they

barely have a phone line. This is a medicalintern in his 20s, who’s extremely com-puter literate, who has keyboard skills,and would love nothing better than tohave the Internet in front of him so he

can see pictures of problems he facesin order to cure people, and use the

Internet to continue his training.” A marketassessment by Era Digital found that only 3

percent of more than 10,000 rural medical clinicsin the Americas had Internet access.

E-Link is banking on DVB-RCS—DigitalVideo Broadcasting Return Channel via Satellite—which was recently published as an ETSI stan-

THINGS TO REMEMBER3E-Link Americas turns the “small-scale struggle” for deployment on its head. However, while its principle aimis broad-scale deployment to leverage economies of scale, at press time it had not lined up a single customeron the ground.3E-Link’s success rests with a long chain of partners, from funders to hardware vendors to local ISPs. Whilepartnerships are extremely important for smooth functioning of any network, for a project of this scale theamount of planning has delayed time to deployment.3E-Link is opting for open-standards technology to push down costs and encourage rapid technologyadvancement.

1.Pages1_End 11/19/03 10:17 AM Page 59


WISPS

dard, to serve a range of potential customers on aregional (not country) basis throughout theAmericas, in order to leverage economies of scale.

DVB-RCS is both cost efficient and scalable. Asimple, single-gateway DVB-RCS system canserve up to thousands of users, while a distributedgateway architecture DVB-RCS system could pro-vide integrated services to several hundred thou-sand users. Customers with small antenna termi-nals (80 to 90 cm in diameter) get high-speedInternet access through high-frequency bands,multi-spot coverage areas, and high-speed digitalsignal processing on board the satellite.

Where connectivity was once achieved on apiecemeal basis (with a single satellite beam), E-Link will connect large numbers of institutions andtelecenters at a rate of 256 Kbps to 2 Mbps for only$75 to $150 per month. It is aiming at generic mar-kets such as health centers (largely scattered in ruralareas), smaller municipalities and institutions out-side major urban areas with no connectivity, tele-centers, and schools. While coverage will be avail-

able throughout a region, connectivity will be pro-vided regionally through local ISP and IT partners.

“What happened in the world is that solutionsto deploy satellites have been proprietary,” Zadrasays. “This is not cost effective and it doesn’t scale.You are held captive to prices and controls thatonly one company can provide. But with open-standards technology, any company that can pro-vide the specifications of the open standards cansupply for us.”

Because it allows hardware manufacturers tofocus on a single technical solution, DVB-RCSmay well become a global satellite standard, pro-viding a healthy and open competitive environ-ment that benefits both industry and users.

“As we’ve seen with Wi-Fi, you get a lot ofhardware producers that push the cost down,”Zadra says.

E-Link ultimately remains not a satellite solu-tion but a wireless solution, bundling both satelliteand Wi-Fi together. It wants to put a Wi-Fi chip inthe satellite so the last mile becomes an 802.11

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 60


WISPS

solution, further lowering costs.

The partner chain

E-Link works through a partnership amongsatellite, hardware, and funding interests as well asincumbent carriers (who can help ensure scalabili-ty further on). Each partner rolls out a differentpiece of the initial deployment.

“If you bundle the capacities of the differentpieces that need to be integrated to provide a solu-tion, then you have a little more cost-effective wayto provide it,” Zadra says. “We have more equip-ment vendors than are required. We have moresatellite proposals than required. It is a good posi-tion to be in because we are trying to determinewhat options provide the best value proposition.”

E-Link is a for-profit company that is essential-ly owned by a nonprofit trust, which disbursesfunds to E-Link which then provides the servicesand takes in revenues. Equipment maintenance isdone by E-Link value-added service providers,which differ from country to country. In eachcase, local partners handle deployments, mainte-

nance, and support.E-link has a threefold strategy to tackle the dif-

ferent regulatory environments among the manycountries of the Americas. It intends to work withwithin-country value-added partners who arealready licensed to provide Internet service.Where regulatory impediments arise, it will askgovernments and ministers for special exemptionsto ensure that, for example, schools can be served.In some instances, it may work to change legisla-tion (though E-Link acknowledges that that thisapproach is time intensive and could take years).E-Link plans to extend service first to those coun-tries with no regulatory restrictions to DVB-IP andWi-Fi, such as Brazil, parts of the Caribbean,Chile, and Mexico.

While actual deployment still lies the future,Zadra says “the issue is getting it right, and not thedemand.” Deployment rests on a network of localcountry-specific partnerships, and E-Link needseverything lined up correctly. It is aiming toward theCaribbean and Central America for its first phase ofdeployment and will move on from there.w

1.Pages1_End 11/19/03 10:17 AM Page 61

1.Pages1_End 11/19/03 10:17 AM Page 62


WISPS

WISP on the edge33UniNet Communications, Cape Town, South Africa

UniNet Communications of Cape Town isan ISP with 11 employees committed toproviding cheaper last-mile connectivity

for home professionals and small businesses inurban areas and to those living in isolation inrural areas. CEO David Jarvis first launched thebusiness in Mozambique, using wireless broad-band to provide a more affordable last-mile solu-tion for customers.

“The service offered by the existing monopoliesin Mozambique and South Africa is of variablequality and very expensive,” Jarvis says. “They arealso often unable or unwilling to cover disadvan-taged areas.”

In Cape Town, a wealthy urban area, UniNet isworking on a phased approach of targeting certainkinds of clients, first to get the company up andrunning and then to expand to rural and disadvan-taged areas later on.

“Unlike most larger companies, we see reallong-term commercial potential in the rural under-

developed sectors of the society,” Jarvis says.“With wireless technologies, we can afford toestablish hotspots in rural areas by focusing oncommunity centers such as local schools.”

Jarvis believes the support for rural and disad-vantaged areas can be done through contracts withgovernment departments or with the support ofNGOs. He said aid organizations are keen to sup-port initiatives that deliver communications intorural areas, providing the initial impetus to estab-lish new services there.

UniNet, for example, has closely investigatedtelecenters run by the University EduardoMondlane in Mozambique. This university has aproactive IT section, and the university runs itsown 802.11 wireless backbone across Maputo city.

The challenge from regulators

Jarvis set up UniNet Lda, the first wireless ISPin Mozambique. In South Africa, UniNet alongwith FrogFoot Networks has designed and manu-

THINGS TO REMEMBER3Taking bold steps to confront regulators could help push needed reforms. Adding international pressure alsohelps. Remember, that extensive openness about your own services can harm your operation and result inequipment confiscation.3At every turn, equipment may need to be modified to meet the needs of the local environment.3As a last-mile solution, wireless broadband networks are cheap and reliable, can be deployed quickly, andexpanded easily.

1.Pages1_End 11/19/03 10:17 AM Page 63


WISPS

factured its own antenna systems, and built itsown Linux-based access concentrators and wire-less repeaters. It has tested a range of wirelessradios and cabling systems and refined itsrequirements to ensure reliability and cost effec-tiveness in a developing environment. And it hasused solar-powered installations because of theunreliable electricity grid and the rural location ofsome of its sites.

In both Mozambique and South Africa, howev-er, UniNet’s activities have threatened statemonopolies, resulting in a series of challenges fromthese countries’ regulatory agencies.

“The big state-owned telecommunicationscompanies are still trying to keep their monopoly,”Jarvis says. “The rules and regulations are not welldefined, and this is generally used to help protectthe existing monopolies.”

In Mozambique, the National Institute ofCommunications, the regulatory authority, told

UniNet it must suspend all services running on 2.4Ghz. But after some legal consultation, UniNetfound there were no regulations for this area ofcommunications services. UniNet resolved torequest formal documentation stating the require-ments for licensing its service, and why it was notlegal to continue. But with no formal documentforthcoming, it decided to continue on.

Jarvis’s experiences in Mozambique gave him theconfidence to tackle the South African market,where the Independent Communications Authorityof South Africa, or ICASA, is far more organizedand active. “The main barrier at the moment is theregulatory authority,” Jarvis says. “A clear under-standing of the legal framework, and the attitudes ofthe local regulatory authority are crucial.”

Cape Town controllers

In July 2002, with an initial investment of$28,000 from its three owners, UniNet began

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 64


WISPS

building a wireless ISP in Cape Town—justenough to establish one repeater on a high site.“From there we expanded and self-financed thesecond repeater site,” Jarvis says. But before long,ICASA called on the tiny start-up.

“They have repeatedly threatened to close usdown and recently confiscated some of our equip-ment from a high site,” Jarvis says.

There are currently five other wireless ISPsoperating in the city. UniNet has 73 clients con-nected to its network, 20 of whom were temporar-ily disconnected because of the confiscation ofequipment at one of its high sites.

In the future, Jarvis says he probably would notexpose UniNet so much. “We were very open withregard to location of our repeaters, the type oftechnology used, and coverage areas. This made it

easier for the regulatory authority to locate ourequipment and facilitated its attempts to investi-gate our services.”

UniNet has presented proposals to otherlicensed telecommunications operators, in thehope of generating interest for a buy-in to help“license” its wireless Internet activities. It is cur-rently awaiting responses to these proposals.UniNet has also been seeking international sup-port for its regulatory difficulties with ICASA. Ithas approached USAID, the World Bank, andlocal government departments in an attempt tolobby support for the deregulation process.

“This process is slowly gaining momentum,”Jarvis says, “but we still often feel very vulnerablein the context of these restrictive regulatoryframeworks.”w

In Cape Town, an ICASA technician disassembles UniNet equipment. In the center photo, CEO David Jarvis tries to pre-vent any damage.

1.Pages1_End 11/19/03 10:17 AM Page 65


WISPS

With the exception of private cor-porate use within certain boundaries,commercial wireless broadband

activity has historically been viewed as illegalwithin the scope of South Africa’s TelecomsAct, according to sources within South Africa’sIndependent Communications Authority.ICASA has no formal position on the legality ofWi-Fi “outdoor” services. Nevertheless, it hasconfiscated equipment on three occasions, dis-rupting the operation of UniNet’s network.

CEO David Jarvis says that ICASA seems tohave withdrawn its legal processes against com-panies like his and backed off from what hecalls its “aggressive attitudes.”

Q. What has ICASA been specifically con-cerned about?

A. We have been running the networks in thelicense-free frequencies, 2.4 Ghz and 5.8 Ghz,but the regulator is unhappy with our existinglicense. ICASA is concerned that we are pro-viding not only a telecommunications servicebut also our own infrastructure across “publicroads”—in other words, we fall into the catego-ry of infrastructure provider.

Q. Describe the telecoms legal landscapeas 2003 draws to a close.

A. At the moment, the framework is chang-ing tremendously. The government is encour-aging a very rapid reform process and intro-duction of a convergence bill in the course ofnext year. Among other things, this law willempower a number of other companies to offerservices similar to those of the state monopoly.

Q. How does wireless broadband fit in?A. Wi-Fi falls into this category to some

extent, but there is vagueness about exactly howthe new bill will interpret its use. There is a moveto limit Wi-Fi to indoor use and/or only on a sin-

gle premises, which unfortunately means thatthen Wi-Fi will be limited to the domain ofdeveloped areas and not developing areas.

Q. What about the competition?A. About 15 companies in South Africa

operate public outdoor Wi-Fi networks, andthere is little collaboration among them,because most of the companies are operatingquasi-underground. In the Western Cape, how-ever, operators have formed a wireless Internetexchange, WINX, which serves as a platformfor collaboration. Frequency planning andstrategies for surviving harassment by the regu-lator are common points of discussion.

Q. Where does UniNet stand in the rangeof wireless broadband providers?

A. We are by no means a major player, butwe’ve had significant interaction with the regu-lator because of confiscations. This is partlybecause we made it easy for the regulator tomake an example of us. We were very public inour advertising. Our Web site is very descrip-tive and informative and open regarding theservices we were providing, and our priceswere competitive.

Q. How exactly did you get into trouble?A. The regulator was encouraged by the state

monopoly to act against us before we made sub-stantial inroads into their market. In SouthAfrica, the regulator does not work on its own.It acts on complaints from an operator. We havereceived threatening letters from the operator,saying they will use legal means to recover lostrevenues if we do not cease operations.

It’s ironic. The municipalities here havewireless networks that connect all their officebuildings, yet the government, through ICASA,is not supportive. In effect, the municipalitiesare fighting the government.w

Going head to head with regulators

1.Pages1_End 11/19/03 10:17 AM Page 66

SHARED ACCESS

1.Pages1_End 11/19/03 10:17 AM Page 67

1.Pages1_End 11/19/03 10:17 AM Page 68


SHARED ACCESS

As rural universities in Africa graduallyacquire Internet access, they can beginbringing social and economic opportunities

to the communities that surround them. The uni-versities can build cyber centers linking (oftenwirelessly) beyond their walls to help pay connec-tivity costs, and the centers themselves can beginto meet local development needs.

Over the past couple of years, the John W.McCormack Graduate School of Policy Studies atthe UMass Boston has partnered with universitiesin Senegal and Kenya to build such university-ori-ented operations.

“Normally, the universities are gated,” saysMargery O’Donnell, Project Administrator with theMcCormack Graduate School. “What this is doingis reorienting the thinking of the universities’ peopleto help surrounding community development.”

McCormack’s community resource centers, or

Scaling Africa’s ivory towers33Partners in Education, Saint Louis, Senegal; Njoro, Kenya

CRCs, are much more than cyber cafés, whichsimply help make the centers self-sustaining.They’re being furnished with business-centerequipment and other components to serve thedevelopment needs of the general population andat-risk youth.

The central idea is to build the CRC, funded byUSAID’s Education for Development andDemocracy Initiative, and then adapt the socialand economic programs to meet the needs of thecommunity and of NGOs interested in working inthe area. The potential for each CRC is virtuallyunlimited and rests with the imagination of thelocal community.

Senegal and Kenya

McCormack’s first partnership was in ruralNorthern Senegal with Université Gaston Bergerde Saint-Louis, whose Internet link enabled

THINGS TO REMEMBER3The McCormack Institute partnerships are primarily interested in bringing social and economic devel-opment opportunities to rural areas. They’ve chosen technology and a wireless solution to achieve thesegoals, but don’t view connectivity alone as a final objective. While the Internet has been critical to pro-viding distance education in Masaailand, in Senegal it has been used to generate returns through acyber café.3The partnerships are both replicable, from Senegal to Kenya, and scalable, within Kenya. 3These deployments take best advantage of an untapped local resource: existing university infrastructure.

1.Pages1_End 11/19/03 10:17 AM Page 69


SHARED ACCESS

McCormack to wirelessly connect a newly con-structed CRC one fourth of a mile from the univer-sity. The CRC has a business center with a faxmachine and a cyber café generating income to sus-tain overall operations. But there’s also a women’scenter, sewing workshop, classrooms, and associat-ed agricultural and transport initiatives.

Because of its success in Senegal, the Institutepushed on to rural Egerton University in Njoro,Kenya, 112 miles northwest of Nairobi. AtEgerton, the deploy-ment began byinstalling a computerlab in the university.(While USAID has con-nected universities inKenya with fiber, theFaculty of Arts andSocial Sciences hadonly one computer.)Here, the CRC is beingbuilt only half a milefrom the gated campuswhere the Internet con-nection to the universityhas been made withfiber. As the CRC isbuilt, the mentoringprograms and onlinedistance learning will beconducted at the uni-versity.

Because of the AIDS crisis in Kenya, the NjoroCRC has a special mission: “Basically, any pro-grams that you can think of relating to AIDS willbe here,” O’Donnell says. This includes testing forthe virus [Voluntary Counseling and Testing(VCTs)], education, mitigation, awareness, andmentoring of young people. Social and economicdevelopment programs will help to remove thestigma and ostracization of those affected byHIV/AIDS, reintegrating them back into the com-munity. Programs will deal with the AIDS orphancrisis (2.1 million in Kenya) and grandmothers

who are the caretakers of those orphans. A dispen-sary and pharmacy are also planned.

The First Lady of Kenya is chairing a boarddeveloped by the Institue that will help to establishthis model countrywide. Long after AIDS hasbeen eradicated in Kenya, the CRCs with their ITcapacity will still be standing.

Maasailand

McCormack is now scaling its Njoro operationto an unservedMaasailand community75 miles to the south ofEgerton University inNarok. The $83,000project will bringInternet-based dis-tance education toNarok’s MaasaiEducation Discovery, afledging CRC workingto serve Maasai youthand to help create eco-nomic developmentopportunities.

The effort here espe-cially targets Maasaigirls, who are deprivedof any education oppor-tunities. “Maybe oneMaasai girl goes to uni-versity each year,”

O’Donnell says. “Most are married young and havemany babies.” About 80 students from Maasailandand Egerton have already been enrolled into dis-tance-education classes to be provided by UMassBoston beginning in January 2004.

Because connecting the Maasai EducationDiscovery has been expensive, the Institute hasbeen debating whether to link it to the Internetusing a satellite or a wireless link back to Egerton.

“The problem with a landline is that it’s justnot going to work in the hinterland of Sub-Saharan Africa,” O’Donnell says. “When the CRC

McCormack project locations in Kenya.

1.Pages1_End 11/19/03 10:17 AM Page 70


SHARED ACCESS

is near the university we can lay the lines, but oncethe CRCs are father away, we need to considerwireless options.”

In October 2003, additional funds were raisedfor the Narok project to purchase a DVB-RCS, orDigital Video Broadcaster and Receiver, for a one-time cost of $10,000. Its wireless antenna has a 7-km radius which can reach computers newlyinstalled in local Maasailand schools. The month-ly usage, however, is $2,000, and the schools areexpected to chip in. “This is an extraordinary stephere in Kenya,” O’Donnell says.

So far, McCormack has not hit any substantialregulatory barriers in Senegal or Kenya. Whileboth countries have been privatizing, “it doesn’t

mean that if we received a billion dollars and wantto put wireless all over Kenya that we will not havea problem,” O’Donnell says. “We’re finding a wayon the ground here. It takes small projects like thisto become important.”

O’Donnell would like to see the spread ofCRCs throughout Kenya’s university system.Kenya has an AIDS mandate, and the First Ladyhas expressed an interest in further expansion ofthe Egerton project. This CRC model in Kenyacould be used elsewhere in Africa to deal with theAIDS pandemic.

“We are trying to break the back of the digitaldivide,” O’Donnell says, “so however we can do it,we will.”w

Plans for Egerton CRC in Kenya. Maasai Education Discovery in Njorok.

The Saint-Louis CRC computer lab.The Saint-Louis CRC in Senegal.

1.Pages1_End 11/19/03 10:17 AM Page 71

1.Pages1_End 11/19/03 10:17 AM Page 72


SHARED ACCESS

Village replication33Sustainable Access in Rural India, Tamil Nadu State, India

In a 2,000-sq-km area that is home to 32,000Indians, Sustainable Access in Rural India, orSARI, has helped 23 percent of the population

connect to the Internet—remarkable, given India’soverall average of 1.5 percent and a worldwideaverage of 9 percent.

In this pilot project, applications and contentare provided wirelessly to more than 80 sites in 50villages in Madurai District, in Tamil Nadu State,which equates to the highest density of ruralInternet kiosks anywhere in the world.

SARI is a public, private, and academic collabo-ration among IIT in Madras, Harvard’s BerkmanCenter for Internet & Society, MIT’s Program onInternet & Telecoms Convergence, and the I-Gyan

Foundation. n-Logue Communications Pvt. Ltd. isthe private-sector implementing partner.

The idea is to tap India’s relatively well-laidfiber infrastructure, which penetrates most towns,and bring low-cost connectivity to surroundingvillages. SARI uses corDECT, a wireless local loopdeveloped by IIT, which currently costs around$320 but should ultimately drop to $200.

n-Logue, the project driver, works togetherentrepreneurially with a start-up local Internetservice provider and village kiosk operators to pro-vide connectivity and services. The LSP in thiscase co-invests with n-Logue to sell connectivity toa radius of about 35 km (generally a couple of smalltowns and about 350 villages). The average size of

THINGS TO REMEMBER3Multiple partners bring a range of perspectives and widen the range of available services.3Users usually need to see a clear benefit to new technology before they will use it. A proactive telekioskoperator can build trust in the technology while pushing applications.3Everyone must benefit. Villagers receive services and kiosk operators and LSPs earn profits—a typicalwin-win situation enabling scaling to and replication in other regions.3The LSP provides valuable structured training for maintenance and applications, helping the kiosk oper-ator buy into the project.3n-Logue’s approach is top down, whereas in Africa the McCormack Institute relies on the community todrive project agenda. At Egerton University, IT is but one tool. In SARI, it is the only tool.3Phones in every home are not affordable, so shared access is fundamental to project start-up andgrowth.

1.Pages1_End 11/19/03 10:17 AM Page 73


SHARED ACCESS

a village with a telekiosk is 1,000 households,while the smallest villages have fewer than 300households. Average per-capita income is under$1 a day.

With a $1,000 loan from a village bank—themaximum amount allowed at that branch level—the kiosk operator invests in a basic kit thatincludes the wireless connection, a multimedia PCwith a color monitor, multimedia equipment andweb camera, a power source with a four-hourback-up battery, and a dot-matrix printer.

The LSP trains the kiosk operators, usuallyyoung men and women from the villages, provid-ing them with software support and technicalskills. The kiosk operators then maintain theequipment and assist customers in sending andretrieving voice and text messages, filling in onlineforms, and using the Web. Above all, these com-

puter-literate early adopters communicate thepotential benefits of the telekiosk to the villagers inan understandable and appealing way.

To break even, the operator must generate $3 perday, charging a minimum fee for each service to payoff the loan and begin making a profit. The operatormust be willing to win the trust of new users, goingthe extra length to create awareness and encouragecomputer usage in order to increase revenues.

The first telekiosks were installed in November2001, and to date, about a quarter are breaking evenor are profitable. Telekiosks in larger villages seemto generate better returns simply because there aremore users there. Early services include communi-cations, education and training, tele-agriculture,tele-medicine, entertainment, and e-government.

Because the infrastructure investments by theLSP are relatively low (averaging about $75,000

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 74


SHARED ACCESS

Because villagers need to be convincedthat paying for Internet services willbenefit them, SARI’s success depends

on telekiosk operators building trust amongusers and pushing a variety of applications(generally tailored for low literacy in English)in five categories: communication, wealth,entertainment, education, and e-governance.Each can generate revenues for the operatorand benefit the villager.

Because many friends and relatives in TamilNadu have emigrated to the Gulf and otherregions in search of jobs, the e-mail, voice, andvideo mail services provided are very useful.Voice and video confer-encing have enabledfamilies separated bygreat distances to com-municate regularly.

In one example, awoman’s kidneys failedafter her husband left forthe Gulf. Hearing this,her in-laws sent herback to her parents, disowning her and sayingshe was defective. This woman managed toget a kidney transplant, but her in-lawsrefused to take her back into the family, andwhen she asked how to contact her husband,they would not give her the information. Shemanaged to obtain his contact through afriend. The husband was sent a letter explain-ing how to obtain an e-mail account and atime was set to videoconference. Sitting thou-sands of miles away, the husband was madeaware of the real situation.

Thanks to e-governance applications, vil-lagers can now send online applications forpensions, birth, death, below-poverty-line, and

encumbrance certificates; register complaintsof broken street lights and drinking-water prob-lems; and send petitions to government politi-cians and bureaucrats, allowing interaction atfive different levels of government.

In Tamil Nadu, the cost to a villager whenapplying for a birth certificate online dropsfrom the usual $6 to $2. A handicapped villagerwho lost his job managing the local water pumpbecause of “local politics” petitioned the chiefminister by e-mail and was subsequently rein-

stated. At least two such petitions are sent eachmonth. A village president who was abusingher position of authority to sell publicly ownedtrees was removed after e-mail complaints fromthe local kiosk operator.

SARI has collaborated with the local agricul-tural college to provide online consultations bye-mail and video conferencing, because smallfarmers can turn only to pesticide suppliers ortheir peers for advice. Some 45 specificinquiries were received from farmers in justfour months through the kiosk’s tele-agricul-ture service. In one case, a farmer avoided los-ing a $2,800 diseased okra crop after receiving,

An infinity of applications

3

1.Pages1_End 11/19/03 10:17 AM Page 75


SHARED ACCESS

per district) and an entire village kiosk set-up costsapproximately $1,000, SARI appears to be highlyscalable. An investment of less than $100,000 perdistrict can be leveraged six or seven times withlocal funding from the LSP and kiosk operators.

Consequently, SARI has been replicated in 7other Indian states through 16 more projects.The current goal is to roll out SARI to 30,000villages over the next two years, and to cover over500,000 villages in the next three to four years,providing access to 85 percent of the ruralIndian population.

Going forward, research teams from IITMadras, MIT, and Harvard University, workingwith a range of local collaborators, will explorenew technologies, including next-generation wire-less systems, new low-cost appliances, and innova-tive application and system software. (Despiteongoing discussions with the national communica-

tions authorities, circuit-switched voice serviceshave not yet been offered.) Social and economicresearch, monitoring, and assessments begununder the Madurai pilot will also continue.w

for only $0.40, treatment advice by e-mail.Most farmers have little access to veterinary

advice other than through unlicensed practi-tioners who often charge exorbitant rates forquestionable diagnoses. Direct access toexperts at the Tamil Nadu Veterinary SciencesUniversity has provided villagers with cures forvarious animal ailments without having to leavetheir village.

In Attapati village, Priya sent a photo of herlimping chicken to the veterinary college. Thedisease was identified as curled-toe paralysiswhich is caused by Vitamin-B deficiency. Themedication was communicated online for a costof only $0.40 versus $4 earlier.

SARI has partnered with Aravind EyeHospital to provide online eye consultations.Kiosk operators are trained take pictures of

eyes and then e-mail them to hospital doctors.Based on these images, doctors can make a pre-liminary diagnosis and recommend variouscourses of action: to make an appointment, trya home remedy, visit a local eye clinic, orundergo an operation.

Using an online questionnaire, villagersthemselves can send digital pictures of theireyes. Diagnoses are e-mailed back to them, andappointments for treatment are made as need-ed. About 80 patients have used this service ina six-month period.

In addition to screening movies, thetelekiosks also provide offline services such asphotography (e.g., for passports) and test train-ing including sample questions and lectures.SARI wants to collaborate with ICICI to pro-vide online banking facilities to villagers andpossibly a low-cost ATM.w

Boys log on at a typical SARI telekiosk, or chiraag, whichmeans “something that brings light” in Hindi.

3

1.Pages1_End 11/19/03 10:17 AM Page 76

ADAPTIVE TECHNOLOGIES

1.Pages1_End 11/19/03 10:17 AM Page 77

1.Pages1_End 11/19/03 10:17 AM Page 78



W ireless broadband can have thebiggest impact in rural areas where thereis the least infrastructure. In these poor

areas, with limited or unavailable connectivityoptions, a little bit of wireless can go a long way.

First Mile Solutions, based in Cambridge,Massachusetts, has been working to develop anaffordable, low-risk wireless infrastructure tobridge the digital divide and jumpstart the ruralcommunications market.

“We don’t really know how people are going touse infrastructure where it hasn’t existed,” saysFounder and Managing Partner Amir AlexanderHasson. “It may not even make sense yet to call ita market, the some 4 billion people living in ruralareas who lack communications infrastructure.”

In collaboration with Professors SandyPentland and Rich Fletcher of MIT Media Lab,Hasson has developed a “store-and-forward”infrastructure for rural areas using local trans-portation. Villages surrounding a town within 10to 80 kilometers have an array of transportation

vehicles passing through on a daily basis, raising apromising connectivity possibility. In addition topicking up and dropping off passengers, what ifyou could do the same with data?

Originally funded and patented by the MITMedia Lab, DakNet—the name comes from theHindi word “dak,” which means “post” or“postal”—has been implemented in India andCambodia (and soon Nigeria). DakNet is essential-ly a store-and-forward wireless broadband net-work using a Mobile Access Point, or MAP, devicethat is mounted on any vehicle that frequentlypasses by a series of villages.

When the vehicle comes in range of a Wi-Fi-enabled kiosk within the village (up to 1 kmdepending on line-of-sight, velocity, and use ofantennas), the MAP automatically senses a wirelessconnection with a kiosk and delivers and collectsdata at an average “goodput” (adjusted for connec-tion-dropping) of 2.3 Mbps. After making a rou-tine circuit, the vehicle returns to a hub, such as acyber café, VSAT, or post office, and uplinks to the

THINGS TO REMEMBER3 Applications that are compatible in different settings increase the chances of their adoption. In India,DakNet worked as an intranet for land record changes. In Cambodia, it is used for e-mail and as a non-real-time search engine.3 A project will run smoothly when every part of the value chain has a buy-in to the project; that is, state gov-ernment, bus driver, end user, and so on.

Off the map33DakNet, Karnataka State, India

1.Pages1_End 11/19/03 10:17 AM Page 79



Internet backbone.The result is broadband intranet wireless access

and applications to areas where there may even beno telephone, for under $500 per village.

Initial installations of DakNet are done by theFirst Mile Solution team, which also buildshuman-resource development capacity by traininglocal teams about wireless networking, empower-ing them to maintain, expand, and upgrade thenetwork themselves.

Although the data transport provided by DakNet

NETWORK SET-UP

A DakNet bus mounted with a MAP. A DakNet hub with satellite uplink.

is not real-time, a very large amount of data can bemoved at once, supporting a wide variety of non-realtime applications, including audio and video mes-saging, e-mail, community bulletin boards, publichealth announcements, and music and video broad-casts. Information-intensive applications include col-lection of environmental sensor information, voting,census/polling, health records, and land records; aswell as Web services such as searching and browsing,e-commerce, and voice-mail over IP. The service canalso track the movement of goods.w

1.Pages1_End 11/19/03 10:17 AM Page 80



In Karnataka State, the local govern-ment has pioneered the computerization ofland records in what is recognized as the

first state-level e-governance initiative in India.The Bhoomi project has been successfullyimplemented at district headquarters acrossKarnataka to completely replace the physicalland-records system.

DakNet has taken Bhoomi one step farther,however, by decentralizing the land-recordsdatabase to villages up to 70 km from the districtheadquarters, or “taluka,” in Doddaballapur.DakNet uses a public government bus with aMobile Access Point, or MAP, to transport land-record requests from each village kiosk back tothe taluka server. The server processes requestsand outputs land records that are then deliveredback to the kiosks for printing and payment(about $0.32 per land record).

Amir Hasson answered questions aboutDakNet’s design and economics and about asecond implementation in Cambodia that deliv-ers Internet access to remote schools through asimilar store-and-forward wireless method.

Q. What did the Karnataka drivers thinkabout having an antenna mounted on theirbuses?

A. All of the bus drivers wanted to be the firstDakNet-enabled bus, but gradually they real-ized that if all these people were doing landrecords from the village, they would not be rid-ing the bus as often, and that meant they willlose revenue. At that point, they could not real-ly object because they had orders fromKarnataka’s revenue department and e-gover-nance department. Nevertheless, to scale thenetwork, you would need to provide the rightincentives for the means of transport as well.

Q. Does DakNet have a business model

Store-and-forward stepping-stonesassociated with Bhoomi?

A. The Bhoomi project really is not designedto make money. It is going to take about 6.5-7years for the total investment in those kiosks topay off if they are doing only land records forRs.15 per record. Other services need to beprovided or we might need to develop cheapmachines that just do land records very cheaply.

Q. Did you develop software for theproject?

A. We worked with the National InformaticCenter and customized their front end to workwith our system. So it uses the same database(backend) but a different front end that worksin a store-and-forward or asynchronous mode.

In India, we just provided an intranet—thebus going back and forth taking land records—but in Cambodia, we are actually providingInternet access. You have full-fledged e-mailaccounts using Outlook Express on each of thecomputers. People have e-mail IDs, and theyare globally connected. With a non-real-timesearch engine, they can also enter a keywordsearch, such as “malaria,” and the search is sentto our server in Cambridge where the cachedresults are e-mailed back to the user who canthen browse the results offline.

Q. What is the business model inCambodia?

A. It is much more of a long-term approachwhere the schools have been funded by theWorld Bank and other donors as communitycenters of excellence with some minimaltuition. But as the schools have an economicdevelopment dynamic within their premises,they may become a commercial center that thevillagers use for their daily needs.

Q. Do you see the store-and-forward

3

1.Pages1_End 11/19/03 10:17 AM Page 81



design as a long-term solution? Is this a step-ping-stone or the end point?

A. Communications is a fundamental rightand there are many advantages to real-time con-nectivity at some point. But what we are argu-ing is that you need a way to get there, andDakNet is a stepping-stone that’s designed tobe more or less seamless. We would like to posi-tion ourselves as a provider of seed technologythat can also grow to provide the technologythat is demanded as each village develops.

A. What are some of the advantages of thestore-and-forward design over a cell phone?

A. There are a lot of disadvantages of provid-

ing a real-time network especially if you’re work-ing in a shared-access kind of model. If you havea telecenter, with one phone booth for the wholevillage, the catch there is that the person who isbeing called is almost never at that phone booth.Someone has to record the message for the per-son who was called and then deliver the mes-sage. If you are paying for that kind of infrastruc-ture, you need to provide access devices(phone/computer/handheld) in households andnot as a shared model. The private sector willperk up when people will start to be able toafford their own access devices and we start tomove away from the shared model. That’s a kindof gold mine here.w

3

1.Pages1_End 11/19/03 10:17 AM Page 82



Life is hard in the villages 85 km north ofVientiane, in rural Laos. Each year bringstorrential rains followed by high tempera-

tures and a thick red dust. The people here aredeprived of telephone connections and power andhave limited access to cellular services because ofthe surrounding mountain range.

But the Jhai Foundation in San Francisco hasbeen busy here introducing techniques to farmersto help them grow surpluses of rice and othercrops. The foundation has shown expert womenweavers how to use natural dyes to weave textilesfor export. To profit from surpluses, however, thefarmers need accurate and timely informationabout pricing in the market town of Phon Hongand in Vientiane. And the women must communi-cate with emigrant Laotians to increase the rate ofreturn from the sale of their woven goods.

After researching alternatives like expensive

THINGS TO REMEMBER3Because off-the-shelf equipment can’t always withstand harsh field conditions, practitioners shoulddetermine whether they will need to research and develop special hardware, including alternative powersources.3Where knowledge of English is low, local-language applications may be critical. When local scripts arenot available and/or literacy is low, voice applications may be an appropriate substitute.3Involving end users in each stage of the project creates a sense of personal ownership and can smoothadoption of the technology.

Beyond ‘off the shelf’33Jhai Remote IT Village Project, Vientiane, Laos

radiotelephones and impractical fiber line, JhaiFoundation’s Remote Village IT program found asustainable, low-cost solution for the villagersthrough the combination of a low-power comput-er, wireless broadband, and voice-over IP.

Overcoming the lack of a power supply andconditions unfriendly to off-the-shelf technologies,Jhai’s team of engineers, led by personal computerpioneer Lee Felsenstein, developed a 12-watt com-puter prototype that can withstand dumping in abucket of water, battering by dust storms, and a fallfrom the table. The team located an Indian-manu-factured, bicycle-pedaled generator to supplypower. (Because of the monsoon season, solar pan-els were too inefficient.) The bicycle cost is onlyone third of a solar-powered set-up. One minute ofpedaling generates about five minutes of power.

The computer connects wirelessly to an anten-na on the roof, which sends a signal to a nearby

1.Pages1_End 11/19/03 10:17 AM Page 83



mountain, which is then forwarded 25 km toPhon Hong, the closest town with phone lines. Awireless LAN centered at the mountaintop relaystation will eventually transmit signals betweenfive villages and a server at Phon Hong Hospital,enabling access to the Internet over the Lao tele-phone system.

The computer will run a Laotian version of theLinux-based graphical desktop KDE, which wasdeveloped by a student-teacher team at theNational University of Laos in Vientiane under thedirection of former IBM software developerAnousak Souphavanh. The suite of business toolsallows Laotians to browse the Web, send e-mail,and create simple documents.

Lee Thorn, Chairman and Founder of the JhaiFoundation, calls the localization into Laotian“probably the hardest part of the whole project.”To generate fonts in Laotian, for example, the teamhad to create its own Unicode, a standard forencoding letters and other characters in any lan-guage. The entire system was built with the needs

of villagers closely in mind. They were involved inevery stage of the project from the very beginning.Minimal use of outside “experts” has encouragedlocal ownership of the project.

This network will enable villagers to make tele-phone calls within Laos and internationally usingvoice-over IP. It will also enable the accounting,letter writing, and e-mail that are so important forthe villagers’ start-up enterprises.

In addition to incubating the business aspirationsof local farmers, builders, and weavers, the set-upwill facilitate communication with distant familymembers who send remittances and who are poten-tial business partners with the villagers. The com-munity will also be able to effectively communicatewith other organizations like the Jhai Foundationand seek development help on its own terms.

“We’re building the system to last for years, awhole social, technical, and economic tool thatlooks like a system and is owned by poor people,”Thorn says. “The Lao members of the team,including the villagers, see this effort as a gift from

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 84



the people of Laos to the world’s rural poor. A sus-tainable, replicable solution.”

While monsoons and road washouts havecaused some delays, testing of the Phon Kham vil-lage link to the Internet began in early 2003. Usinga laptop, Vorasone Dengkayaphichith, Jhai’s coun-try coordinator, sent e-mail and pictures, and, onJanuary 23rd, used VoIP to speak to colleagues inCanada, the United States, Sweden, and Laos. Thesignal went from Phon Kham village via a station ina tree on a hill (2.4 km) to a water tower in PhonHong (9 km) and then to the Internet via a phoneline. Software problems foiled a first attempt toinstall in February, and the team plans to return toLaos in November 2003 to implement this project.

Although the Jhai project north of Vientianestill faces certain logistical issues, it has receivedinquires from India, Indonesia, and many othercountries where telecommunications and powerdistribution remain limited. The project teamplans to refine the Jhai approach for use inother locations.w

SSoolliidd ssttaatteeJhai Foundation’s Remote Village IT programfound a sustainable, low-costsolution for villagers with its own durable, theft-resistant computer. TheJhai Computer, with an MZ-104 CPU board, 64MB of RAM, and a 96 MBdisk-on-chip, is designed to withstand harsh con-ditions, including dust,dirt, water, and humidity.

A bicycle-powered computer in Vientiane.A Jhai team member at the “Wi-Fi Tree.”

A Lao-enabled browser.

PHOT

O BY

LIA

M H

ELM

ER

1.Pages1_End 11/19/03 10:17 AM Page 85

1.Pages1_End 11/19/03 10:17 AM Page 86

REBUILDING NATIONS

1.Pages1_End 11/19/03 10:17 AM Page 87

1.Pages1_End 11/19/03 10:17 AM Page 88


REBUILDING NATIONS

Internet Project Kosovo, or IPKO,launched wireless broadband connectivity inKosovo on September 20, 1999, in the weeks

after the war, as a project of the InternationalRescue Committee (IRC). NATO bombing haddestroyed much of Kosovo’s communi-cations infrastructure. The Serbs whoran it were gone.

At the time, UN and HumanitarianAgencies had no effective way of com-municating or sharing information, andthey were clamoring for connectivity.Most were using expensive satellite-based communications, and the defaultwas for each to deploy its own solution. But thiswas expensive and duplicative and did not connectthe local community. IPKO’s alternative was toinstall one satellite dish and build an 802.11 wire-less local loop network around it.

The deployment kicked off with a $175,000loan from IRC and a 3.8-mm satellite dish donated

by Interpacket Networks. The team installed thedish and an 802.11b wireless network in Pristina.IPKO sold service to every UN agency, NATO,international NGOs, Internet cafés, and business-es. It provided free connectivity to schools, hospi-

tals, the university, and local NGOs.“We were in a rather blessedly unreg-

ulated environment,” remembers IPKOcofounder Paul Meyer. “Our license wasa half-page M.O.U. with the UN. Thatserved as our operating license for threeand a half years until finally the govern-ment passed a telecom law. So I can singthe virtues of free regulation. It really

was the Wild West.”After six months, IPKO was incorporated as a

local NGO. In the course of its first year, it beganexpanding to other main urban centers in Kosovo.Nine months after inception, it launched a dial-upservice. After a year of operation, it opened theIPKO Institute of Technology, a technology train-

Wi-Fi Wild West33Internet Project Kosovo, Pristina, Kosovo

THINGS TO REMEMBER3Initial financial support is critical. The funding can be a donation, a financial loan, or loan of equipment.3Open regulatory environments and zero bureaucratic hassles expedite the deployment of wireless tech-nologies.3Connectivity for its own sake will not ensure project adoption. Look beyond the technology and developapplications to meet the needs of local users.

1.Pages1_End 11/19/03 10:17 AM Page 89


REBUILDING NATIONS

An IPKO dish in Pristina.

ing institute in Pristina that now provides CISCO,Microsoft, and other training courses to more than200 students at a time.

“IPKO was founded to provide the tools,knowledge, and environment required for Kosovoto participate in the global information society,”Meyer says. “It has done so.”

In late 2000, its Internet-service-provider busi-ness was spun off as a for-profit subsidiary(IPKONet) owned by the nonprofit Institute and itsemployees. IPKONet now has a nationwide 155-MBwireless backbone, a wireless link across Albania tothe fiber-optic backbone in the Adriatic Sea, Wi-Finetworks in the main cities of Kosovo, and nationaldial-up access. It is the leading Internet provider andone of the largest businesses in Kosovo.

IPKO was profitable from the outset, and itsstart-up loan was repaid in four months. In theyears since its founding, it has raised commercialbank debt to finance network upgrades andexpansion.

IPKO attributes much of its success to Kosovo’spermissive regulatory environment and the critical

mass of organization customers. Both the satellitedish donation and a working capital loan were cru-cial to kicking off, while a commitment to handover the system met with good hands in well-trained Kosovar managers and technicians. Today,Meyer recommends relying on local infrastructureand local operators where they exist, providingaccess to the community and civil society, andcommitting to leaving the system behind.w

NETWORK SET-UP

1.Pages1_End 11/19/03 10:17 AM Page 90


REBUILDING NATIONS

At the “Wireless Internet Opportunity forDeveloping Countries” conference at UNHeadquarters on June 26, 2003, Paul

Meyer, Co-Founder of Internet Project Kosovoand CEO of Voxiva, urged the development com-munity deploying ICTs to “focus on the informa-tion flow problems you’re trying to solve, not onthe technology.”

“I think that International development agen-cies have spent too much time thinking abouttechnologies and networks and not enough timethinking about information flow problems.Technology is not an end in itself. It is only valu-able if it facilitates flows of key information andaddresses critical problems. Start with the fun-damental problem: What information needs toflow where? Then figure out how to put togetherexisting technologies, networks, and applicationsto solve it.

“Connectivity is not enough. The potential ofinformation technology is not realized when kidscan chat and download music videos at Internetcafés in the developing world. The potential isrealized when people at the periphery can partici-pate meaningfully in the broader world; whenbusinesses can transact seamlessly with one anoth-er; when governments can make decisions informedby accurate, timely information and provide criti-cal services to their citizens.

“There are three lessons I’ve taken from myexperience at IPKO in Kosovo and at Voxiva, thecompany I now run.

“First, international-development agenciesshould not be telecom companies. They shouldwork with and become customers of existingproviders. At IPKO in Kosovo, we built a nation-al wireless backbone because we had to. There wasno existing infrastructure. Luckily, most coun-tries in the world have existing and rapidlyexpanding telecom, mobile phone, and Internetnetworks. Go to the far corners of Uganda, and

there’s mobile telephony. Companies like MTN inUganda and IPKO in Kosovo are building outinfrastructure. International development agen-cies and policy makers should focus on the policyproblems, the information flow, and the applica-tions and solutions that are required to promotehealth or education, for example. That focus onapplications does much more good in promotingaccess because it creates a market and demand forconnectivity. By focusing on the policy problemsthey are chartered to solve, international develop-ment institutions can be aggregators of demandfor connectivity.

“The second lesson I’ve learned is that applica-tions are harder than infrastructure. Figuring outwhat key pieces of information need to flow whereis hard work. For example, designing and imple-menting a decision support and communicationapplication for a Ministry of Health to track dis-ease outbreaks requires deep understanding of thework processes of the doctors and administrators atevery level. Furthermore, introducing a system ofreal-time information necessarily changes the wayan organization functions, making change man-agement and training keys to success. People whoare experts in health or education or microfinanceshould focus their energies on figuring out howthose sectors can benefit from real-time informa-tion flows. They shouldn’t spend their time worry-ing about connectivity.

“The final lesson I’ve learned is that there are alot more phones in the world than computers.Moreover, phones don’t require literacy. If thedeveloping world is to benefit from applicationsthat address information flow problems in key sec-tors like health and education, we need to lookbeyond just the Internet and PCs. If applicationscan be extended to every computer and every phonein the developing world, then we will have thereach to make a real difference in solving criticalproblems there.”w

Applications are harder than networks

1.Pages1_End 11/19/03 10:17 AM Page 91

1.Pages1_End 11/19/03 10:17 AM Page 92


REBUILDING NATIONS

Today, only 6 percent of the populationof Afghanistan has access to electricity.Many Afghans have never made a phone

call, let alone sent e-mail or surfed the Internet.The nation is arguably at the far end—the bottomend—of the digital divide.

But the UN Development Program inAfghanistan is working to change this, at least for asmall but growing percentage of Kabul’s 2.5 mil-lion residents, thanks to a $300,000 grant fromFrance, a wireless LAN, and new telekiosks at theairport and in post offices less than 10 km from theMinistry of Communications.

Four of the telekiosks, which include five PCsand a printer, were opened on July 6, 2003. Twowere linked to the Ministry with fiber, and, by mid-August, two were connected wirelessly to base sta-tions at the Ministry. Each post-office telekioskcosts $19,500 to set up. If revenue predictions arecorrect, each is expected to generate up to $30 perday from local users.

In the first month, the initial four telekiosks

received 2,600 customer visits. In the secondmonth, with six telekiosks in operation, there were2,410 customer visits. The number droppedbecause of power-supply problems and becausetwo of the six telekiosks were only open for half amonth. UNDP reports that all nine telekiosks werefunctioning by late September.

Two “animators” at each telekiosk helpAfghans learn basic computer and Internet skills.While most users are male, an outreach effort isunder way to bring in more women and girls. Thetelekiosks should enable schoolteachers and chil-dren, government employees, health workers, andothers in Kabul to do e-mail and access theInternet. A Web site will support both Dari andEnglish and eventually Pashto.

The majority of the people visiting thetelekiosks are first-time users and need basiccomputer training. “We had anticipated that mostpeople visiting the telekiosks would be returningrefugees from places where they received com-puter training or workers at international organi-

Picking up the PCs33UN Development Program, Kabul, Afghanistan

THINGS TO REMEMBER3In a nation or region recovering from war or crisis, wireless networks can serve as a fast substitute fornational telecom infrastructure, especially where there are few regulatory barriers.3Making use of existing solid structures for shared Internet access can reduce the cost of deployment.3Sporadic availability of power can shut down operations, causing revenue losses.

1.Pages1_End 11/19/03 10:17 AM Page 93


REBUILDING NATIONS

zations that would want to use the Internet,” saysRafiqullah Kakar, the National ProjectCoordinator for UNDP-Afghanistan. “But mostusers were first-time users of technology.” A feefor training may be added; $.20 -.40 per hour atone telekiosk could generate another $5 to $10per day.

In Kabul, infrastructure is still limited, soUNDP’s decision to link the telekiosks with awireless LAN was made early. “It was a cost-effec-tive way to provide a quick and easy connectionto the Ministry’s satellite Internet backboneaccess while avoiding any infrastructure prob-lems that may face a landline implementation,”reports Karen Gray, a project consultant toUNDP, in a memorandum.

The benefits of wireless LAN outnumber theinconvenience of installing masts up to 120 feettall in some locations. The WLAN is high band-width and quick and easy to install. It faces noregulatory issues, can be self-managed, and canlink to the Ministry’s 1.5 Mbps satellite Internet

backbone.Three base stations were installed at the

Ministry to give 360-degree coverage of the citywithin a range of 10 km. The amplifier has 250 mwof transmit power. The farthest telekiosk from theMinistry is 8 km. Each base station unit can sup-port 64 sites (total 194 sites) because each is set toa different channel (within the allowed frequencyband). Currently, there is only sufficient band-width on the satellite data access backbone to sup-port 20 sites.

As the UNDP telekiosks are expanded toAfghanistan’s other provinces, the network designwill need to be revised. Some provinces have oneto five post offices in the main city. The more ruralprovinces have one post office in the main city andothers spread as much as 50 to 100 km from themain post office.

A specific technology goal for the deployment isto determine whether a wireless LAN really is fea-sible in such a challenging environment. Toencourage sustainability, UNDP has committed to

Most telekiosk users in Kabul have been male, but UNDP is reaching out to women and girls.

1.Pages1_End 11/19/03 10:17 AM Page 94


REBUILDING NATIONS

Avid newcomers to the Web.

providing hands-on and classroom training onwireless networking to Ministry employeesinvolved in the project.w

NETWORK SET-UP

Kabul’s airport telekiosk.

1.Pages1_End 11/19/03 10:17 AM Page 95

1.Pages1_End 11/19/03 10:17 AM Page 96

GUIDELINES TO COUNTRIES

1.Pages1_End 11/19/03 10:17 AM Page 97

1.Pages1_End 11/19/03 10:17 AM Page 98

GUIDELINES TO COUNTRIES 99

W ireless Internet technologies andapplications are in their early develop-mental stages, and their potential eco-

nomic and social benefits appear to be consider-able especially in the context of the UnitedNations Millennium Development Goals.International development experts and leadingIT corporations consider wireless Internet tech-nologies essential to bridging the digital divide indeveloping countries at a manageable cost andwithin a reasonable timeframe.

These new technologies and applications, how-ever, are emerging in many different situations,often outside the operational landscape of tradi-tional telecommunications services, and with newtypes of participants from both the private andpublic sectors. Wireless Internet applicationsresulting from grassroots initiatives may be seen asdisruptive and can hit unintended roadblocks inthe form of local regulations and lack of under-standing of their potential. Conversely, proper gov-ernment support and incentives can acceleratetheir successful implementation at little cost andwith significant immediate economic and socialbenefits for the poor.

Each country presents unique characteristicsand conditions with respect to wireless Internetdeployment, from a geographic, social, economic,regulatory, and telecommunications infrastructurestandpoint. There are, however, several regulatoryand economic success factors that country author-ities should consider. To facilitate such localreview, the following guidelines to governmentswanting to support the experimentation and earlydeployment of wireless Internet infrastructureapplications, are proposed.

1. Identify, promote, and establish national con-sensus on the potential benefits of wirelessInternet applications and local priorities.

A key factor of success for wireless Internetdeployment is a generally supportive environment,a prerequisite of which is adequate awareness ofpolicy makers, the public and private sectors, andthe local media.

United Nations Millennium Development GoalNo. 8, Target 18, states: “In cooperation with theprivate sector, make available the benefits of newtechnologies, especially information and commu-nications.” There is a general agreement that

Proposed Guidelines in Support of LocalExperimentation and Early Deployment ofBroadband Wireless Internet

Prepared by the Secretariat of the United Nations ICT Task Force and the WirelessInternet Institute on the basis of discussions at the “Wireless Internet Opportunityfor Developing Countries” conference at United Nations Headquarters in New YorkCity on June 26, 2003.

1.Pages1_End 11/19/03 10:17 AM Page 99


Internet access is a key success factor in makingthese new technologies work for development. Inmany geographic areas, however, the cost ofexpanding Internet connectivity with landlines orcable remains prohibitive for the foreseeablefuture, while available bandwidth is increasingexponentially in developed countries, furtheringthe digital divide.

Raising awareness and building national con-sensus about the benefits of low-cost broadbandwireless Internet infrastructure solutions is there-fore an important first step.

One complex challenge is to clearly differentiatethe broadband capabilities of IEEE 802.xx-basedtechnologies for fixed infrastructure from the low-bandwidth mobile solutions offered by current cel-lular telephony and personal-devices networks.

Local governments may also inventory thoseapplications that may contribute the most to bridg-ing the digital divide both from a geographic andsocial standpoint, and foster economic develop-ment, job creation, and productivity gains in alleconomic sectors. Of particular interest for devel-oping countries are opportunities in the 3 “e’s”: e-government, e-education, e-health (a fourth, e-com-merce, is sometimes also included); rural-areas cov-erage, small business connectivity, the developmentof local ICT services including private-sector ISPsand voice-over IP services.

Identifying leading applications, which maydrive the initial use of wireless Internet infra-structure and distribution, will further developsupport for wireless Internet solutions among keyconstituents.

2. Adopt minimum regulations supportingthe use of unlicensed spectrum and ICT indus-try standards.

A second critical success factor is the leverageof internationally recognized norms.Harmonization creates significant economies ofscale for equipment manufacturers, softwaredevelopers, and off-the-shelf solutions. It is there-fore in the best interest of countries to closely fol-low and support the norms and recommenda-tions of the following organizations:

3For spectrum policy and standards references:3Institute of Electrical and Electronics

Engineers (IEEE)3ITU World Radiocommunication Conference3For manufacturers standards compliance, inter-operability, and quality assurance:3Wi-Fi Alliance for 802.11 products3WiMax Forum for 802.16 productsThe single-most important step to support wire-

less Internet applications is to develop a spectrumpolicy that allocates bands for unlicensed applica-tions, such as the 2.4- and 5-GHz bands currentlyunlicensed in a number of countries.

In a first step toward worldwide spectrum allo-cation to wireless Internet applications, the ITU inits July 4, 2003 World Radio Conference commu-niqué indicates that it “successfully establishednew frequency allocations to the mobile service inthe bands 5150-5350 MHz and 5470-5725 MHzfor the implementation of wireless access systemsincluding RLANs. Wireless devices that do notrequire individual licenses are being used to createbroadband networks in homes, offices, andschools. These networks are also used in publicfacilities in so-called hotspots such as airports,cafes, hotels, hospitals, train stations, and confer-ence sites to offer broadband access to theInternet....

“The lower part of the 5-GHz spectrum will bepredominantly used for indoor applications withthe first 100 MHz (5150-5250 MHz) restricted toindoor use. The use of these frequency bands isconditional to provisions that provide for interfer-ence mitigation mechanisms and power-emissionlimits to avoid interference into other radio com-munication services operating in the same spec-trum range.”

Each country may have existing spectrum allo-cations that need to be adjusted. A close monitor-ing of international regulatory activities and theadaptation of local spectrum policies is thereforerecommended.

3. Update telecommunications regulations tofoster market opportunities, optimize existinginfrastructure resources, and free competition

1.Pages1_End 11/19/03 10:17 AM Page 100

GUIDELINES TO COUNTRIES 101

among wireless Internet service providers.Although situations will vary greatly from one

country to another, four areas need careful attentionfrom a telecommunication regulatory perspective:

(a) Access to an Internet backbone is a prereq-uisite for successful deployment of broadbandwireless Internet services. To the extent possible,telecommunications regulations regime should fos-ter a competitive Internet backbone market.Private- or public-sector entities outside thetelecommunications industry should be allowed tocontribute their backbone capacity. This is partic-ularly important in regions with underused back-bone capacity. As an example, GrameenPhone inBangladesh was made possible by the collabora-tion of the National Railway System contributingaccess to its fiber-optic network. Conversely, ifonly a few backbone service providers exist, theyshould be obligated to open up their infrastructureto independent service providers.

(b) If no backbone is available, alternative solu-tions include broadband terrestrial wireless links,satellite stations and power grids. Appropriateincentives encouraging such solutions should beimplemented.

(c) At the final distribution level, Internet serviceproviders, or ISPs, implementing wireless Internetsolutions need to be facilitated by telecommunica-tions regulations and operate in the context of freemarket competition.

(d) Legacy regulations concerning radio com-munications, some related to law-enforcementagencies or the military, can also create unintend-ed obstacles for wireless Internet ISPs includingfor the physical deployment of antennas andradio-communications equipment. These shouldbe adapted to the specific needs of wireless ISPsas in such landmark cooperation agreements facil-itated by the regulator between the private sectorand the NTIA (National Telecommunicationsand Information Administration) on use of unli-censed spectrum in the United States (seeAppendix 1 on p. 103).

4. Identify key available resources and fostercooperation among potential actors.

Wireless Internet infrastructure and servicescan leverage a number of existing resources in anygiven country:3backbone operators and owners of fiber-opticnetworks including governments, private-sectornetworks, telcos, power-grid operators, and satel-lite communications operators;3owners of real estate and high points with ade-quate power supply and security to install anten-nas, such as existing radio communications towersor possibly public-sector buildings such as postoffices or other types of standard venues;3systems integrators with the technical capabili-ties to install and maintain wireless equipmentsuch as towers, cabling, hub-integrated antennas,wireless modems, control and network manage-ment systems, routers, cables, uninterruptiblepower supplies, racks, etc.;3operators of similar services such as TV broad-casting, cellular telephony, computer maintenanceorganizations, and power distribution;3financial-sector interested to fund start-up ISPs;3community leaders eager to provide Internetaccess to their constituents;3incumbent telecommunications operators; and3providers of know-how and training services.

It is recommended to support venues such as con-ferences or seminars on wireless Internet to increaseawareness and initiate dialogues and cooperationamong key decision makers from both the privateand public sectors.

5. Support the experimentation of new servicesand encourage the aggregation of demand forbandwidth.

Although wireless Internet initiatives can beperceived as disruptive and infringing on exist-ing regulations, they are an important source ofinformation to adapt regulations and developpublic-sector strategies. Governments areencouraged to support early initiatives withinterim measures until appropriate regulatoryadjustments are made.

One of the most important factors of success forwireless ISPs is a rapid increase of initial demandfor connectivity, which allows for a faster break-

1.Pages1_End 11/19/03 10:17 AM Page 101


even on operating expenses. This can be achievedthrough initial aggregation of demand based onapplications for local public services such asschools, universities, health services, and publicadministration. Business, agriculture, and privateuses will inevitably add to the mix once service isavailable. In very underserved areas, it is likely thatinitial viable aggregation will occur through thedeployment of wireless Internet kiosks operatedby small entrepreneurs.

Governments are encouraged to deploy applica-tions in the area of e-government, e-education, ande-health to leverage the use of Internet access by thegeneral public. Governments should also encourage

local public services to use the infrastructure of localwireless ISPs.

6. Follow-up and support of wireless Internetdevelopments at governmental and intergov-ernmental levels, including sharing of bestpractices.

It is anticipated that wireless Internet tech-nologies and applications will continue to evolverapidly, which makes it important for govern-ments and private-sector leaders to remainabreast of other countries’ experiences, regulatorywork at the international level, best practices,and latest innovations.w

1.Pages1_End 11/19/03 10:17 AM Page 102

APPENDIX 1 103

National Telecommunications and Information AdministrationFor Immediate Release January 31, 2003

Contact: Clyde Ensslin or Ranjit de Silva, 202-482-7002

AGREEMENT REACHED REGARDING U.S. POSITION ON 5-GHzWIRELESS ACCESS DEVICES

Executive Branch agencies of the U.S. government, in cooperation with the FederalCommunications Commission (FCC) and industry proponents of 5 GHz Wireless AccessSystem (WAS) devices, have reached agreement on modifications to the U.S. position withrespect to these devices, for use in international fora.

“I am very pleased that the participants from the federal government and the privatesector have come to a consensus that satisfies both of their interests,” said AssistantSecretary of Commerce and NTIA Administrator Nancy J. Victory. “In achieving this,the Bush Administration has continued its goal of stimulating the economy and ensur-ing the national defense and preserving the leadership of the U.S. high-tech sector.Based on these changes, the U.S. is now able to formalize its position with respect toearth exploration satellite systems, mobile, and radiolocation services at 5 GHz and willnow fully support these allocations,” Victory said.

The NTIA, FCC, NASA and Department of Defense (DoD), working closely withindustry in detailed technical meetings, have agreed to modify the required DynamicFrequency Selection (a listen-before-transmit mechanism) detection threshold charac-teristics contained in the U.S. proposal for WRC-03 Agenda Item 1.5 as follows:

A. The Threshold value is modified from -67 dBm to -64 dBm for 1W to 200 mWdevices and -62 dBm for < 200 mW devices. This would now apply to the 5250-5350MHz and 5470-5725 MHz bands.

B. The U.S. position of only allocating the 5150-5350 MHz band to the mobile serv-ice will be modified to include an allocation to the mobile service in the 5470-5725MHz band and the resolution for continued studies on this band will be deleted.

C. As a consequence, of these modifications, the U.S. position on the EarthExploration Satellite Service allocation at 5460-5570 MHz will be to support thisallocation.

The changes to the U.S. position will now allow the U.S. to seek a mobile allocationat the International Telecommunication Union (ITU) World RadiocommunicationConference in Geneva in June and July 2003 in both bands sought by industry (5150-5350 MHz and 5470-5725 MHz) while ensuring protection of vital DoD radars.

The previous technical requirements for these systems, which must include DFS,have also been modified based on new information recently explored. These changesnot only continue to ensure that the vital DoD radars are protected, but ease sharingconditions for the WAS systems.

1.Pages1_End 11/19/03 10:17 AM Page 103

1.Pages1_End 11/19/03 10:17 AM Page 104

VENDOR RESOURCES

1.Pages1_End 11/19/03 10:17 AM Page 105

1.Pages1_End 11/19/03 10:17 AM Page 106

VENDOR RESOURCES 107

By Amir Alexander Hasson

The IEEE 802.11- and 802.16-based wirelessnetworking technologies—collectively, thewireless Internet—have emerged as a digital-

communications standard over the past four years,producing a dynamic value chain of suppliers, ven-dors, and consumers. The technology’s affordabil-ity, ease of setup, standardization, and favorableregulatory environment in target markets haveenabled rapid development, with some clear costand performance advantages over wire-line net-working (i.e., Ethernet) for homes and offices. Ithas reduced total cost of ownership for networkingand served as a high-performance medium to dis-tribute available bandwidth.

Although the market for wireless Internet tech-nologies has become saturated in some areas andfailed to meet some bullish industry projections, ithas been one of the only large growth areas withinthe IT and telecommunications sectors during therecent economic slump, and many strategists con-sider it a disruptive technology.

The wireless Internet value chain is comprisedof three broad segments, beginning with networkequipment manufacturers, network software andapplication providers, and network integrators andoperators (see Figure 1).

Equipment manufacturers produce the basicnetworking components that constitute the physi-

cal layer of any wireless network. As networkequipment developed, so did the need for tools tomanage, operate, and protect wireless networks,which led to several new companies entering thevalue chain as application and software providers.

The variety of these tools, combined with theincreasing demand for wireless networks in publicand enterprise IT markets, led to an increasingneed for packaged solutions, spawning anothergroup of new entrants (and new divisions withinconsulting companies) known as system integra-tors. Once a critical installed base of wirelessInternet-enabled devices and users was estab-lished, new market opportunities to capitalize onand operate wireless networks as an Internetaccess service emerged. This led to the widespreadproliferation of public and private wirelessInternet-enabled networks throughout the UnitedStates, Europe, and parts of Asia-Pacific.

The equipment manufacturer segment of thevalue chain consists of chipset makers, accesspoint and client radio producers, antenna andamplifier vendors, and producers of wireless-enabled access devices (Figure 2).

Chipset makers create the core wireless technol-ogy radios and firmware (software that enableswireless networking to work with operating sys-tems and other hardware components). Thesechipsets are then integrated into wireless hardware

Understanding the wireless Internetvalue chain*

*The listings in this chapter are not meant to be exhaustive but indicative of available resources in each of the categories.

WWiirreelleessss IInntteerrnneettEEqquuiippmmeennttMMaannuuffaaccttuurreerrss

WWiirreelleessss IInntteerrnneettAApppplliiccaattiioonn aannddSSooffttwwaarree PPrroovviiddeerrss

WWiirreelleessss IInntteerrnneettIInntteeggrraattoorrss aannddOOppeerraattoorrss

EENNDD UUSSEERRSSpp pp pp

FFiigguurree 11:: AAnn oovveerrvviieeww ooff tthhee wwiirreelleessss IInntteerrnneett vvaalluuee cchhaaiinn..

1.Pages1_End 11/19/03 10:17 AM Page 107


products such as client radio cards (used for lap-tops and PDAs) and access points, which connectclient radios to an available backbone such asDSL. Other companies focus on producing anten-nas and amplifiers required to connect wirelessclients with access points over long distances or inunfavorable networking environments (802.16technology). As the demand for wireless network-ing products has grown, manufacturers of PCs,PDAs, and even cell phones have also begun tointroduce 802.11-compliant wireless chipsets intotheir products. This segment of equipment manu-facturers has seen tremendous growth in recentyears and has been the primary winner within theoverall wireless Internet value chain.

The application and software provider segmentconsists of providers for network security; networkoptimization; network management; and integra-

tion with back-end systems for accounting, billing,and customer-relationship management (Figure 3).

One of the caveats of wireless networks hasbeen their vulnerability to hackers, prompting awide range of providers to develop security appli-cations designed specifically for wireless Internet.This is perhaps the most saturated segment withinthe entire value chain with dozens of competingproviders. Other companies have recognized aneed for software that makes wireless networks(and the engineers who deploy then) “smarter,”including network testing and optimization tools,advanced routing and power-management proto-cols, and network training programs.

Perhaps the most interesting feature within thissegment, into which significant research has beeninvested, is “mesh networking” which aims toenable any 802.11-enabled device to also serve asa router to connect surrounding 802.11-enableddevices to the Internet indirectly through otherdevices that may be connected to the Internet.Another important area within this segment is net-

WWiirreelleessss CChhiippsseettss

ProximIntersilAgereAironet (Cisco)IntelAtherosTexas InstrumentBroadcom

AAcccceessss PPooiinnttss aannddCClliieenntt RRaaddiiooss

AlvarionNomadixValuePointLinksys (Cisco)D-LinkProximSymbolSMC3COMNetGearMotorolaFiretideLucentAlcatelSiemensHughes

22..44//55..33 GGhhzzAAnntteennnnaass aanndd AAmmpplliiffiieerrss

Hyperlink Tech.CushcraftBuffaloHughesLuxul Corp.Pacific WirelessFractusMotiaVivato

WWiirreelleessss--EEnnaabblleedd AAcccceessss DDeevviicceess

DellIBMHP-CompaqMotorolaToshibaNokiaSiemensPalmSony

pp pp pp

FFiigguurree 22:: TThhee wwiirreelleessss nneettwwoorrkkiinngg eeqquuiippmmeenntt sseeggmmeenntt..

Amir Alexander Hasson is Founder and ManagingPartner of First Mile Solutions in Cambridge,Massachusetts.

1.Pages1_End 11/19/03 10:17 AM Page 108

VENDOR RESOURCES 109

work management software, which enables net-work administrators to monitor and administratewireless networks alongside other existing net-works through interfaces that capture critical vari-ables relating to network usage and performance.

NNeettwwoorrkk SSeeccuurriittyy

BlueSocketElixarEscalon NetworksFortress TechnologiesFunk SoftwareICSA LabsNetwork ChemistryNewbury NetworksSenforceTropos

NNeettwwoorrkk OOppttiimmiizzaattiioonn

Berkeley VaritronicsSystemsMesh NetworksRadiant NetworksekahauWild PacketsCWNP ProgramTropos Networks

NNeettwwoorrkk MMaannaaggeemmeenntt

Birdstep TechBridgewater SystemsCirond TechNetNearUAirwaveSyBaseWiDeFi

AAccccoouunnttiinngg,,BBiilllliinngg,, CCRRMM

iPassBilling ConceptsRockSteadyhereUare Comm.Service FactoryExcilanPeoplesoftNomadixBirdstep TechGRIC

pp pp pp

FFiigguurree 33:: TThhee wwiirreelleessss IInntteerrnneett aapppplliiccaattiioonn aanndd ssooffttwwaarree sseeggmmeenntt..

Finally there has been a particular need to devel-op applications that enable wireless Internet serv-ice operators to monetize network usage and tienetwork management software into other enter-prise and legacy systems critical to business oper-

IInntteeggrraattiioonn aannddDDeeppllooyymmeenntt

Concourse CommMake ‘M WirelessBelAir NetworksFirst Mile SolutionsNortel NetworksIBMPronto NetworksWiproWavelinkWayportAlcatelSiemens

PPrriivvaattee HHoottssppoottOOppeerraattoorrss

HotspotzziPassT-MobileMobileStarTelia-SoneraWayportSurf and SipWiFi MetroSkyNetGlobalRoomLinxBoingo WirelessGRIC CommHotel ChainsRestaurants

(Starbucks)Airports (San Jose)

PPuubblliicc HHoottssppoottOOppeerraattoorrss

BAWUG (SanFran)NYC WirelessConsume (London)BC Wireless

(Canada)Madrid WirelessXnet Wireless

(Australia)Wireless FranceStockholmOpen.netPrenzel.net (Berlin)Hotel Chains

(Marriott)Guerilla.net

IInndduussttrryyOOrrggaanniizzaattiioonnss

FierceWirelessWiFi PlanetWiFinderW2IWiMAXWiFi AllianceWECAACMIEEEITU World Radio

ConferenceMuni Wireless

pp pp pp

FFiigguurree 44:: TThhee wwiirreelleessss IInntteerrnneett iinntteeggrraattoorrss aanndd ooppeerraattoorrss sseeggmmeenntt..

1.Pages1_End 11/19/03 10:17 AM Page 109


ation. Many companies within this segment of thevalue chain are likely to be consolidated into largerIT consulting firms and system integratorsbecause of over-saturation (Figure 4).

With the multitude of wireless networkinghardware and software products on the market,entrepreneurs and larger telecom and IT consult-ing companies have created custom solutions tobring wireless Internet to vertical markets such ashealthcare, hospitality, utilities, real estate, retail,warehousing, field service and sales, and last-milecommunications. In addition to these verticalmarkets and home and office networking, wire-less Internet has also been deployed to operatepublic (free) and private (commercial) networks,or “hotspots.” Private hotspot operators havestruggled to find a successful business model,which has caused some very ambitious operatorsto fail to meet expectations. This results partlyfrom other providers (ranging from homes torestaurants with Internet connections and wire-less Internet access points) offering wirelessInternet access for free. As Scott Rafer, Founderof WiFinder, suggests, “wireless Internet is likeair conditioning,” meaning that there is a largepopulation of consumers who have grown toexpect wireless Internet access as part of a service

offering. In this sense, wireless Internet accessprovides businesses such as hotel chains and cor-porate offices with reduced networking costs anda differentiation factor for marketing, but it maynot yield increasing returns as a new revenuestream. The entire value chain has also been driv-en and reinforced by the wireless Internet indus-try and standards bodies that have promoted andadapted wireless technologies according to thebest interest of key stakeholders.

Although parts of the wireless Internet valuechain have become saturated and are undergoingconsolidation, it has produced tremendous valuefor consumers through rapidly decreasing costsof wireless products as well as the widespreadavailability of wireless Internet access. The pri-mary winners in this value chain so far have beenequipment manufacturers and end users. Theplayers and business models in the middle of thevalue chain are less mature and more fragmented.Although wireless Internet markets in the UnitedStates and Europe have become increasingly sat-urated from the supply side, there exist manyopportunities for wireless Internet in emergingmarkets in Latin America, Asia, and Africa wherethere is even a greater need for affordable, dis-tributive communications technology.w

1.Pages1_End 11/19/03 10:17 AM Page 110

111

3COM350 Campus Dr.Marlborough, MA 01752-3064Tel: 1-800-638-3266Fax: 508-323-1111Web: 3com.com

ACM/Assoc. Computing Machinery1515 BroadwayNew York, NY 10036Tel: 212-626-0500Web: acm.org

AgereLehigh Valley Central Campus1110 American Parkway NEAllentown, PA 18109Tel: 1-800-372-2447E-mail: [email protected]

Aironet (Cisco)Cisco Systems, Inc.170 West Tasman Dr.San Jose, CA 95134Tel: 408-526-4000Tel: 800-553-6387Web: cisco.com

AirWave Wireless Inc.1700 S. El Camino RealSte. 500San Mateo, CA 94402Tel: 650-286-6100Fax: [email protected]

Alcatel54, rue La Boétie75008 ParisFranceWeb: alcatel.com

AlvarionInternational Corporate Hq.21a HaBarzel St.POBox 13139Tel Aviv, 61131IsraelTel: 972-3-645-6262Fax: 972-3-645-6222E-mail: [email protected]

Atheros Communications, Inc.529 Almanor Ave.Sunnyvale, CA 94085-3512Tel: 408-773-5200Fax: [email protected]

BelAir Networks Inc.603 March Rd.Kanata, OntarioCanada K2K 2M5Tel: 703-736-8306E-mail: [email protected]

Berkeley Varitronics Systems, Inc.Liberty Corporate Park255 Liberty St.Metuchen, NJ 08840Tel: 732-548-3737Fax 732-548-3404Web: bvsystems.com

Billing Concepts, Inc.7411 John Smith Dr., Ste. 200San Antonio, TX 78229Tel: 888-393-5854Fax: 210-692-0720E-mail: [email protected]: billingconcepts.com

Birdstep Technology ASABryggegata 7N-0250 OsloNorwayTel: 47-24-13-47-00Fax: 47-24-13-47-01E-mail: [email protected]

BlueSocket7 New England Executive Park, 4th Flr.Burlington, MA 01803Toll free: 866-633-3358, Press 3Europe: 44-0-1256-475-744E-mail: [email protected]

Boingo Wireless™, Inc.1601 Cloverfield, Ste. 570Santa Monica, CA 90404Sales and Service: 1-800-880-4117Fax: 1-800-880-4117E-mail: [email protected]

Vendor directory

1.Pages1_End 11/19/03 10:17 AM Page 111


DellOne Dell WayRound Rock TX 78682Tel: (512) 338-4400Fax: (512) 728-4238Web: dell.com

D-Link2F, NO. 233-2 Pao-Chiao Rd.Hsin-Tien, TaipeiTaiwan, Republic of ChinaTel: 886-2-2916-1600Fax: 886-2-2914-6299Web: dlink.com

Ekahau, Inc.12930 Saratoga Ave., Ste. B-9Saratoga, CA 95070Tel: 1-866-4EKAHAUFax: 1-408-725-8405E-mail: [email protected]

Elixar, Inc2 Davis Dr.POBox 13169RTP, NC 27709Tel: 1.888.435-4927Web: elixar.com

Escalon Networks9229 Klemetson Dr., Ste. 200Pleasanton, CA 94588Tel: 925-425-9601Fax: 925-425-9603E-mail: [email protected]

Excilan51 Ave. de la GareL-1611 LuxembourgTel: 352-26-12-75-1Fax: 352-26-12-75-60Web: excilan.com

Fierce Wireless1319 F St. NW, Ste. 604Washington, DC 20004Tel: 202-628-8778Web: fiercewireless.com

Firetide, Inc.928 Nuuanu Ave., Ste. 200Honolulu, HI 96817Tel: 808-528-0007E-mail: [email protected]

Bridgewater Systems303 Terry Fox Dr.Ste. 100Ottawa, OntarioCanada K2K 3J1Tel: 866-297-4636E:mail: [email protected]

Broadcom Corporation16215 Alton ParkwayPOBox 57013Irvine, California 92619Tel: 949-450-8700Fax: 949-450-8710Web: broadcom.com

Buffalo Technology (UK) Ltd176 Buckingham Ave.Slough, Berkshire, SL1 4RDUnited KingdomTel: 44-0-1753-55-50-00Fax: 44-0-1753-53-54-20E-mail: [email protected]

Cirond Technologies1999 South Bascom Ave.Ste. 700Campbell, CA 95008Tel: 866-824-7662E-mail: [email protected]

Concourse Communications39 S. LaSalle St. Ste. 1424Chicago, IL 60603Tel: 312-357-2900Fax: 312-357-2959Web: concoursecommunications.com

Cushcraft Corporation48 Perimeter Rd.Manchester, NH 03103Tel: 603-627-7877Fax: 603-627-1764E-mail: [email protected]

CWNP ProgramPO Box 20063Atlanta, GA 30325Tel: 404-305-0555.E-mail: [email protected]

1.Pages1_End 11/19/03 10:17 AM Page 112

VENDOR DIRECTORY 113

First Mile Solutions, LLC432 Columbia St., Ste. B13BCambridge, MA 02141Tel: 617-494-1001Fax: 616-494-6006Web: firstmilesolutions.com

Fortress Technologies4025 Tampa Rd., Ste. 1111Oldsmar, FL 34677Tel: 813-288-7388Web: fortresstech.com

FractusAvda. Alcalde Barnils, 64-68Edificio Testa-Módulo CSant Cugat del Vallés08190 Barcelona (Spain)Tel: 34-935-442-690Fax.: 34-935-442-691E-mail: [email protected]

Funk Software222 Third St.Cambridge, MA 02142Tel: 617-497-6339/800-828-4146Fax: 617-547-1031E-mail: [email protected]

GRIC Communications, Inc1421 McCarthy Blvd.Milpitas, California 95035Tel: 408-955-1920Fax: 408-955-1968E-mail: [email protected]

hereUare Communications, Inc.1600 Adams Dr.Menlo Park, CA, 94025Tel: 650-688-5778Fax: 650-688-5889Web: hereuare.com

Hotspotzz Inc.265 East 100 South, Ste. 245Salt Lake City, UT 84111Tel: 801-415-8090E-mail: [email protected]

Hewlett-Packard3000 Hanover St.Palo Alto, CA 94304-1185Tel: 650-857-1501Fax: 650-857-5518Web: hp.com

Hyperlink Tech.1201 Clint Moore Rd.Boca Raton FL 33487Tel: 561-995-2256Fax: 561-995-2432E-mail: [email protected]

Hughes Network Systems11717 Exploration Ln.Germantown, MD 20876Tel: 301-428-5500Fax: 301-428-1868Web: hns.com

IBM Corporation1133 Westchester Ave.White Plains, NY 10604Tel: 1-800-IBM-4YOUWeb: ibm.com

IEEE3 Park Ave., 17th Flr.New York, NY 10016-5997Tel: 212-419-7900Fax: 212-752-4929Web: ieee.org

Intel Corporation2200 Mission College Blvd.Santa Clara, CA 95052Fax: 408-765-9904Web: intel.com

Intersil Corporation Headquarters675 Trade Zone BlvdMilpitas, CA 95035Tel: 408-935-4300FAX: 408-945-9305Web: intersil.com

iPass3800 Bridge ParkwayRedwood Shores, CA 94065Tel: 650-232-4100Fax: 650-232-4111Web: ipass.com

ITUPlace des NationsCH-1211 Geneva 20SwitzerlandTel: 41-22-730-51-11E-mail: [email protected]

1.Pages1_End 11/19/03 10:17 AM Page 113


Linksys (Cisco)17401 Armstrong Ave.Irvine, CA 92614Tel: 800-546-5797Web: linksys.com

Lucent Technologies600 Mountain Ave.Murray Hill, NJ 07974-0636Tel: 908-508-8080, Press 9E-mail: [email protected]

Luxul Corporation905 North Main St., Ste. D1North Salt Lake, UT 84054Tel: 801-299-0999 Ext. 121Fax: 801-299-0911E-mail: [email protected]

Mesh NetworksPO Box 948133Maitland, FL 32794Tel: 407-659-5300Fax: 407-659-5301E-mail: [email protected]

Motia Inc.225 Lake Ave., Ste. 710Pasadena, CA 91101Tel: 626-405-4435Fax: 626-405-4436E-mail: [email protected]

Motorola2900 South Diablo WayTempe, AZ 85282,Tel: 800-759-1107Web: motorola.com

[email protected]

NetGear Inc.4500 Great America Pkwy.Santa Clara, California 95054Tel: 408-907-8000Fax: 408-907-8097Web: netgear.com

NetNearU2908 Finfeather Rd.Bryan, TX 77801Tel: 877-599-7253Fax: 979-775-4393E-mail: [email protected]

Network Chemistry1170 University Ave.Palo Alto, CA, 94301Tel: 650-575-1425Fax: 253-323-9244Web: networkchemistry.com

Newbury Networks, Inc.745 Boylston St.Boston, MA 02116Tel: 617-867-7007Fax: 617-867-7001E-mail: [email protected]

NokiaNokia Head OfficeKeilalahdentie 2-4, PO Box 226FIN-00045 Nokia GroupFinlandTel: 358-7180-08000Web: nokia.com

Nomadix31355 Agoura Rd.Westlake Village, CA 91361Tel: 818-597-1500E-mail: [email protected]

Nortel Networks8200 Dixie Rd., Ste. 100Brampton, OntarioL6T 5P6, CanadaTel: 905-863-0000Web: nortelnetworks.com

NYC Wireless532 LaGuardia Place, Ste. 138New York, NY 10012E-mail: [email protected]

Pacific Wireless693 E. Draper Heights Wy.Ste. 210Draper, UT 84020Tel: 801-572-3024Fax: 801-572-3025Web: pacwireless.com

Palm400 N. McCarthy Blvd.Milpitas, CA 95035Tel: 408-503-7000Fax: 408-503-2750Web: palm.com

1.Pages1_End 11/19/03 10:17 AM Page 114

VENDOR DIRECTORY 115

Peoplesoft4460 Hacienda Dr.Pleasanton, CA 94588-8618Tel: 800-380-7638Web: peoplesoft.com

Pronto Networks4637 Chabot Dr, Ste. 350Pleasanton, CA 94588Tel: 925-227-5500Fax: 925-460-8227E-mail: [email protected]

Proxim935 Stewart Dr.Sunnyvale, CA 94085Toll free: 800-229-1630Tel: 408-731-2700Fax: 408-731-3675Web: proxim.com

Radiant NetworksThe Mansion, Chesterford ParkLittle Chesterford, EssexCB10 1XL EnglandTel: 44-0-1799-533-600Fax: 44-0-1799-533-601E-mail: [email protected]

Rocksteady Networks3410 Far West Blvd. Ste. 210Austin, Texas 78731Tel: 512-275-0571Fax: 512-275-0575E-mail: [email protected]

RoomLinx6553 Via SerenoRanch Murieta, CA 95683Tel: 800-576-1655Fax: 916-354-0544E-mail: [email protected]

Service FactoryÅrstaängsvägen 17, SE-117 43Stockholm, SwedenTel: 46-8-18-00-26Fax: 46-8-18-00-29E-mail: [email protected]

Siemens AGWittelsbacherplatz 2D-80333 MunichGermanyTel.: 49-89-636-00Fax: 49-89-636-52-000Web: siemens.com

SkyNetGlobal Ltd.Ste. 16, 123 Liverpool St.Sydney NSW 2000AustraliaTel: 612-8251-3800Fax: 612-8251-3809E-mail: [email protected]

SMC38 TeslaIrvine, CA 92618Tel: 800-762-4968Fax: 949-679-1481Web: smc.com

SONY6-7-35 KitashinagawaShinagawa-kuTokyo 141-0001JapanWeb: sony.net

Surf and Sip, Inc.470 Third St., Ste. 100San Francisco, CA 94107Tel: 415-974-6321Fax: 415-777-2469E-mail: [email protected]

SyBaseOne Sybase Dr.Dublin, CA 94568Tel: 925-236-5000Web: sybase.com

Symbol Technologies, Inc.One Symbol PlazaHoltsville, NY 11742-1300Tel: 866-416-8545Fax: 631-738-5990E-mail: [email protected]

1.Pages1_End 11/19/03 10:17 AM Page 115


Telia-Sonera ABMårbackagatan 11SE-123 86 FarstaSwedenTel: 46-0-8-504-550-00Fax: 46-0-8-504-550-01E-mail: [email protected]

Texas Instruments, Inc.Product Information Center13532 N. Central ExpresswayM/S 3807Dallas, Texas 75243-1108Tel: 972-644-5580Fax : 972-927-6377Web: ti.com

T-MobilePO Box 37380Albuquerque, NM 87176-7380Tel: 800-937-8997Web: t-mobile.com

Toshiba1-1, Shibaura 1-chome, Minato-ku,Tokyo 105-8001 JapanTel: 81-3-3457-4511Fax : 81-3-3456-1631Web: toshiba.com

Tropos Networks1710 S. Amphlett Blvd., Ste. 304San Mateo, CA 94402Tel: 650-286-4250Fax: 650-286-4259E-mail: [email protected]

Vivato139 Townsend St., Ste. 200San Francisco, CA 94107Tel: 415-495-1111Fax: 415-495-6430Web: vivato.net

Wavelink11332 NE 122nd Way, Ste. 300Kirkland, Washington 98034Tel: 1-888-699-9283Fax: (425) 823-0143E-mail: [email protected]

Wayport8303 N. Mopac, Ste. A-300Austin, TX 78759Tel: 972-791-3900E-mail: [email protected]

WiDeFi Inc.476 A1A, Ste. 3Satellite Beach, FL 32937Tel: 321-777-2085E-mail: [email protected]

Wi-Fi Alliance2716 Barton Creek Blvd, Ste. 2024Austin, TX 78735Tel: 512-347-7660E-mail: [email protected]

WiFi PlanetWeb: wi-fiplanet.com

WiFinderE-mail: [email protected]: wifinder.com

Wild Packets Inc.1340 Treat Blvd, Ste. 500Walnut Creek, CA 94597Tel: 925-937-3200E-mail: [email protected]

WiMAX ForumWeb: wimaxforum.org

Wipro TechnologiesDoddakannelli Sarjapur Rd.Bangalore-560 035IndiaTel: 91-80-8440011E-mail: [email protected]

Wireless Internet Institute225 Franklin St., 26th Flr.Boston, MA 02110Tel: 617-439-5400Fax: 617-439-5415E-mail: [email protected]: w2i.org

Xnet WirelessPO Box 555 Mornington 3931AustraliaTel: 03-5976-2185E-mail: [email protected]

1.Pages1_End 11/19/03 10:17 AM Page 116

117

Recommended readings

802.11 Wireless Networks: The Definitive Guide (O’Reilly Networking), byMatthew S. Gast. O’Reilly & Associates; 1st edition (April 2002). ISBN: 0596001835.

Book of Wi-Fi, The: Install, Configure, and Use 802.11b Wireless Networking, byJohn Ross. No Starch Press; (February 2003). ISBN: 188641145X

Brave New Women of Asia: How Distance Education Changed Their Lives, editedby Kanwar and Taplin. This report reviews the impact of distance education on womenin Hong Kong, India, Pakistan, Sri Lanka, and Bangladesh. The report encourageswomen to participate in open and distance learning.

Bridging the Digital Divide: Gyandoot: The Model for Community Networks, byRajesh Rajora (2002). Tata McGraw-Hill, New Delhi. This book covers theoreticalmodels of community access, networking infrastructure, information needs assessment,user behavior, research findings and recommendations for similar projects.

Build Your Own Wi-Fi Network, by Shelly Brisbin, Glen Carty (Contributor).McGraw-Hill Osborne Media; (October 22, 2002). ISBN: 0072226242

Building Wireless Community Networks, by Rob Flickenger. O’Reilly & Associates;1st edition (December 15, 2001). ISBN: 0596002041.

China Dawn: The Story of Technology and Business Revolution, by David Sheff(2002). Documents and investigates the Internet sector in China, as well as India, Japanand Singapore.

CWNA Certified Wireless Network Administrator Official Study Guide (ExamPW0-100), by Devin Akin. McGraw-Hill Osborne Media; 2nd edition (February 18,2003). ISBN: 0072229020.

CWSP Certified Wireless Security Professional Official Study Guide (Exam PW0-200), by Planet3 Wireless. McGraw-Hill Osborne Media; (August 2003). ISBN:0072230126.

Deploying License-Free Wireless Wide-Area Networks, by Jack Unger. Cisco Press;1st edition (February 26, 2003). ISBN: 1587050692

Field Guide to Wireless LANs for Administrators and Power Users, by ThomasMaufer. Prentice Hall PTR; 1st edition (October 17, 2003). ISBN: 0131014064.

1.Pages1_End 11/19/03 10:17 AM Page 117


First Mile of Connectivity: Advancing Communications for Rural Development,edited by Richardson and Paisley. FAO, (1998). This book is divided into four sections:Lessons Learned from other Media and Participatory Communications Practices;Starting with the First Mile of Connectivity: Affordable, Adaptable, Profitable andPractical Rural Telephone and Telecommunications Systems; ICTs as Tools to SupportParticipatory Communication Initiatives for Rural and Agricultural KnowledgeSystems; The Policy Context: The Keystone to Improving Rural Telecommunicationsand ICT Application for Development. The book is available from the website of theUN Food and Agricultural Organization.

Fixed Broadband Wireless System Design, by Harry R. Anderson. John Wiley &Sons; (February 27, 2003). ISBN: 0470844388

Maximum Wireless Security, by Cyrus Peikari and Seth Fogie. SAMS; Book and CD-ROM edition (December 18, 2002). ISBN: 0672324881

Public Wi-Fi LAN Threat Becomes Wireless Carrier Opportunity, by YankeeGroup. MarketResearch.com; ISBN: B00007KSFP; (October 1, 2002).

Voice-over IP and IP Telephony: Referenceshttp://www.cis.ohio-state.edu/~jain/refs/ref_voip.htm

Wi-Fi Handbook: Building 802.11b Wireless Networks, by Konrad Roeder, FrankD., Jr. Ohrtman. McGraw-Hill Professional; (April 10, 2003). ISBN: 0071412514.

1.Pages1_End 11/19/03 10:17 AM Page 118

119

3G—The third generation of mobile communications specified by the ITU promises tooffer increased bandwidth and high-speed data applications up to 2 Mbps. It works overwireless air interfaces such as GSM, TDMA, and CDMA.

802.11 Standard—The series of wireless standards developed by the IEEE.Commonly known as Wi-Fi.

802.11a—A wireless networking specification, assigned by IEEE, in the 5-Ghz fre-quency range with a bandwidth of 54 Mbps.

802.11b—A wireless networking specification, assigned by IEEE, in the 2.4-Ghz fre-quency range with a bandwidth of 11 Mbps.

802.11g—A wireless networking specification, assigned by IEEE, in the 2.4 Ghz fre-quency range with a bandwidth of 54 Mbps.

802.16—A group of broadband wireless communications standards for metropolitanarea networks (MANs) developed by a working group of the IEEE.

802.20—A specification of physical and medium access control layers of an air interfacefor interoperable mobile broadband wireless access systems, operating in licensedbands below 3.5 Ghz, optimized for IP-data transport, with peak data rates per user inexcess of 1 Mbps. It supports various vehicular mobility classes up to 250 km/h in aMAN environment and targets spectral efficiencies, sustained user data rates, and num-bers of active users that are all significantly higher than achieved by existing mobile sys-tems. This standard is under development.

Access Point—A wireless hardware device connected to a wired network that enableswireless devices to connect to a wired LAN.

Ad-Hoc Mode—See Peer-to-Peer.

Analog—Modulated radio signals that enable transfer of information such as voice anddata.

Antenna—A device used for transmitting and/or receiving radio signals, whose shapeand size is determined by the frequency of signal it is receiving.

Authentication—A process of identifying a user, usually based on a username and pass-word, ensuring that the individual is whom he or she claims to be, without saying any-thing about the access rights of the individual.

Glossary

1.Pages1_End 11/19/03 10:17 AM Page 119


Backbone—The central part of a large network to which two or more subnetworks link.It is the primary path for data transmission. A network can have a wired backbone or awireless backbone.

Bandwidth—The amount of data a network can carry, i.e., how much and how fast dataflows on a given transmission path. It is measured by bits or bytes per second.

Base Station—The central radio transmitter/receiver that maintains communicationswith mobile radiotelephone sets within a given range (typically a cell site).

Bits per second (bps)—The number of bits that can be sent or received per second overa communication line.

Bluetooth Wireless Technology—A short-range wireless specification that allows forradio connections (2.4 Ghz) transmitting voice and data between devices (such asportable computers, personal digital assistants, or PDAs, and mobile phones) within a30-foot range of each other.

Broadband—A fast Internet connection generally above 200 Kbps. However, no officialspeed definition exists for broadband services.

Cable—A broadband transmission technology using coaxial cable or fiber-optic linesthat was first used for TV and is now being used for Internet access.

Channel—A path along which a communications signal is transmitted.

Client Devices—These communicate with hub devices such as access points and gate-ways. They include PC cards that slide into laptop computers, PCMCIA modulesembedded in laptop computers, and mobile computing devices.

Client—An end user, i.e., any computer connected to a network that requests services(files, print capability), from another member of the network.

Consumer Premise Equipment (CPE)—Devices located at home or office such as tele-phones, PBXs, and other communication devices.

Dial-Up—A communication connection using standard copper wire telephone net-work.

Digital—Technology used in telecommunications where information is processed byfirst converting it to a stream of ones and zeros, permitting extremely complicated sys-tems to be designed and manufactured at reasonable cost through the use of ASICs andcomputer circuitry while meeting very high performance standards.

DNS—A program that accesses a database on a collection of Internet servers to translateURLs to Internet packet (IP) addresses.

1.Pages1_End 11/19/03 10:17 AM Page 120

121GLOSSARY

DSL—Various technology protocols for high-speed data, voice, and video transmission overordinary twisted-pair copper POTS (Plain Old Telephone Service) telephone wires.

Ethernet—Also called 10Base T, Ethernet is an international standard for wired networks. It canoffer a bandwidth of about 10 Mbps and up to 100 Mbps.

Firewall—Software, hardware, or a combination of the two that prevents unrestricted access intoor out of a network.

Gateway—A combination of a software program and piece of hardware that passes data betweennetworks. In wireless networking, gateways can also serve as security and authentication devices,access points, and more.

Hertz (Hz)—The unit for expressing frequency (f ), a measure of electromagnetic energy. OneHertz equals one cycle per second.

Hotspot—A place where users can access Wi-Fi service for free or a fee.

Hotzone—An area where users can access Wi-Fi service free or for a fee.

Hub—A multiport device used to connect several PCs to a network.

IEEE—Institute of Electrical and Electronics Engineers (www.ieee.org), in New York, is anorganization composed of engineers, scientists, and students and is best known for developingstandards for the computer and electronics industry. In particular, the IEEE 802.xx standards forlocal-area networks are widely followed.

Internet Appliance—A computer intended primarily for Internet access, generally offering cus-tomized Web browsing, touch-screen navigation, e-mail services, entertainment, and personalinformation management applications.

Internet Protocol (IP)—A set of rules used to send and receive messages at the Internet addresslevel.

IP Address—A 32-bit number that identifies each sender or receiver of information that is sentacross the Internet. An IP address has two parts—an identifier of a particular network on theInternet and an identifier of the particular device (which can be a server or a workstation) with-in that network.

ISS—A special software application that allows all PCs on a network access to the Internet simul-taneously through a single connection and Internet Service Provider (ISP) account.

Local Area Network (LAN)—A high-speed network that connects a limited number of com-puters in a small area, generally a building or a couple of buildings.

1.Pages1_End 11/19/03 10:17 AM Page 121


MAC—A unique identifier that can be used to provide security for wireless networks.All Wi-Fi devices have an individual MAC address hard-coded into it.

Mapping—Assigning a PC to a shared drive or printer port on a network.

Network Interface Card (NIC)—A type of PC adapter card that works without wires(Wi-Fi) or attaches to a network cable to provide two-way communication between thecomputer and network devices such as a hub or switch.

PC Card—A credit-card-sized removable peripheral that plugs into a special slot onportable computers (and some desktop models), including Wi-Fi cards, memory cards,modems, NICs, hard drives, etc.

Peer-to-Peer Network (P2P)—Also known as Ad-hoc mode, a network of computersthat has no server or central hub. Each computer acts both as a client and network serv-er. It can be either wireless or wired.

Personal Area Network (PAN)—A casual, close-proximity network where connectionsare made on the fly and temporarily. Meeting attendees, for example, can connect theirBluetooth-enabled notebook computers to share data across a conference-room table,but they break the connection once the meeting is over.

Personal Digital Assistant (PDA)—is a digital handheld device that is can transmitdata and services such as paging, data messaging, computing, telephone/fax, email, etc.possible.

Plug and Play—A feature of a computer system which enables automatic configurationof add-ons and peripheral devices like wireless PC Cards, printers, scanners and multi-media devices.

POTS—Standard analog telephone service.

Radio Frequency (RF)—Any frequency within the electromagnetic spectrum associat-ed with radio-wave propagation.

Range—The distance a wireless signal can reach.

Repeater—A device that receives a radio signal, amplifies it, and retransmits it in a newdirection. Repeaters are used in wireless networks to extend the range of base-stationsignals, thereby expanding coverage—within limits—more economically than by build-ing additional base stations.

Roaming—The ability to move from one access point coverage area to another withoutlosing connectivity.

1.Pages1_End 11/19/03 10:17 AM Page 122

123GLOSSARY

Satellite Broadband—A wireless high-speed Internet connection provided by satel-lites. Some satellite broadband connections are two-way—up and down. Others areone-way, with the satellite providing a high-speed downlink and then using a dial-uptelephone connection or other land-based system for the uplink to the Internet.

Server—A computer that lets other computers and devices on a network share itsresources. Includes print servers, Internet servers, and data servers. A server can also becombined with a hub or router.

Software Defined Radio (SDR)—Refers to wireless communication in which the trans-mitter modulation is generated or defined by a computer, and the receiver uses a com-puter to recover the signal intelligence.

Spectrum Allocation—The range of frequencies designated by a NationalTelecommunications Regulatory Authority for a category of use or uses.

Switch—A network device that selects the path that a data packet will take to its nextdestination ensuring optimal network performance. The switch opens and closes theelectrical circuit to determine whether and to where data will flow.

TCP/IP—The underlying technology behind the Internet and communicationsbetween computers in a network. The first part, TCP, is the transport part, whichmatches the size of the messages on both end and guarantees that the correct messagehas been received. The IP part is the user’s computer address on a network.

Unlicensed Spectrum—The government sets up general rules, such as the power lim-its on devices, and then allows any device that meets those standards to operate (unli-censed) in that spectrum.

Very Small Aperture Terminal (VSAT)—An earthbound station used in communica-tions of data, voice, and video signals, excluding broadcast television, consisting of twoparts: a transceiver placed outdoors in direct line-of-sight to the satellite, and a deviceplaced indoors to interface the transceiver with the end user’s communications device,such as a PC.

Virtual Private Network (VPN)—A private network of computers at least partially con-nected by public phone lines. An example is a private office LAN that allows users to log inremotely over the Internet (an open public system). VPNs use encryption and secure proto-cols such as PPTP to ensure that unauthorized parties do not intercept data transmissions.

Voice-Over IP (VoIP)—Technology that supports voice transmission via IP-basedLANs, WANs, and the Internet.

Wide Area Network (WAN)—A network that connects computers and other devicesacross a large local, regional, national, or international area.

1.Pages1_End 11/19/03 10:17 AM Page 123


Wi-Fi Alliance—A coalition of wireless-industry leaders committed to the open inter-operability of 802.11 IEEE standards.

WiMax Forum—A coalition of wireless-industry leaders committed to the open inter-operability of all products used for broadband wireless access based on 802.16 IEEEstandards.

Wireless Internet Service Provider (WISP)—An organization providing wirelessaccess to the Internet.

Wireless Loop (WL)—A wireless system providing the “last mile” connectivity; that is,the last wired connection between the telephone exchange and the subscriber’s tele-phone set (which can be up to several miles in length). Traditionally, this has been pro-vided by a copper-wire connection.

Wireless—Use of radio-frequency spectrum to transmit and receive voice, data, andvideo signals for communications.

World Radiocommunication Conference (WRC)—An international conferenceorganized by ITU at which standards and interference issues are discussed at the inter-governmental level.

1.Pages1_End 11/19/03 10:17 AM Page 124

125

2.4 Ghz, xiii, xiv, 7-9, 17, 18, 25,64, 66, 100, 108

2G, 113G, xiv, 10, 11, 184G, xiv, 115-6 Ghz, 66, 108802.11, xiii, 7-14, 25, 26, 29,

32, 60, 63, 89, 100, 107, 108802.16, xiii, xiv, 9-14, 22, 100,

107, 108access points, xiv, 7-14, 17, 25,

79-82, 108, 110Adriatic Sea, 90Afghanistan, 93-95Africa, xiii, 4, 19, 27, 53-56, 63-

71, 73, 110Africa-Online, 53-56AIDS, 69-71Aironet, 43airport, 8, 49, 93, 95, 100, 109Albania, 90Algeria, 56Alvarion, 54, 57Amazon Association, 47-50Amazon, 47-50amplifier, 38, 94, 107, 108Anatel, 48antenna, xiv, 9-13, 25, 38, 60,

64, 71, 79, 81, 101, 107, 108Aravind Eye Hospital, 76Asia, 3Asian Development Bank, 28asynchronous service, xiv, 26, 81backhaul, 9, 56Baja Wireless, 57, 58Bangladesh, xv, 1, 27-29, 101Barker coding, 8base station, 10-12, 94Berkman Center for Internet &Society, 73Bhoomi Project, 79-82

Botswana, 56Brazil, xiii, 1, 47-50, 61building, 10, 11, 14, 43, 66, 101buy-in, xvi, 38, 65, 79Cable, 7, 17, 43, 56, 100, 101Caboclo Indians, 47-50California, 56Cambodia, 79-82Caribbean, 61cellular network, 28; service,

xiii, 83; standards, xiv; technology, 2; telephony, xiii, xiv, 1,4, 10, 11, 14, 63, 101

Central America, 61Centrino, 9, 21, 22, 108Chad, 19Chile, 61Cisco, 21, 22, 43, 90Commonwealth Development, 28Community Resource Center,

69-71competition, xiv, xvi, 10, 17, 18,

29, 32, 33, 66, 100, 101complementary code keying, 8convergence, xiv, 66, 73copper, xiv, 11, 25, 26, 31, 38,

53corDECT, 73-76DakNet, 26, 79-82digital divide, v, xi, xiii, xv, 34,

37, 79, 93, 99, 100, 117D-Link, 57, 58DSL, 7, 56, 108DSSS, 8DVB, 59-61, 71Eduardo Mondlane, University,

63Education for Development andDemocracy Initiative, 69Egerton University, 69-71Egypt, 19, 56

E-Link Americas, 59-61Ensenada, 57, 58Era Digital, 59Ernest Kleinwort CharitableTrust, 48Ethernet, 7, 11, 107Europe, xiv, 17-19, 25, 53, 107,

110FCC, xiv, 20-22, 103fiber, xiii, 10-13, 25, 31, 38, 43,

53, 70, 73, 83, 90, 93, 101First Mile Solutions, 79-82, 108fixed line, xiii, 2, 4, 57France, 89FrogFoot Networks, 63Gabon, 56Gangetic Plain Project, 25, 26Gaston Berger de Saint-Louis,

Université, 69-71Geneva, 21, 103Gonofone, 27-29GPRS, 11Grameen Bank, 28, 29GrameenPhone, xv, 1, 27-29,

101Guinea, 56Half-Way Proposal, 56harmonization, 100Harvard University, 76Hewlett-Packard, 21Himalayas, 43hotspot, xiii, xiv, 17, 63, 100,

110hotzone, 57, 58ICASA, 63-66ICICI, 76IEEE, xiii, xiv, 7-9, 11, 25, 26,

100, 107, 109I-Gyan Foundation, 73Independent CommunicationsAuthority of South Africa, 63-66

Index

1.Pages1_End 11/19/03 10:18 AM Page 125


India, 19, 25, 26, 73-76, 79-83,85

Indian Institute of Technology,25, 73, 76

Indonesia, 85InfoDev, vii, ix, xiInformation society, v, xi, 25, 37,

90Institute for Connectivity of theAmericas, 59International Rescue Committee,

89Internet Project Kosovo, 89-91Internet service providers, xv,

17, 18, 31, 37, 53-66, 101Interpacket Networks, 89IPKO, 89-91Iraq, 1IRC, 89ISM, xiii, 7, 55ISM, xiii, 7, 55ITU, xiv, 8, 11, 100, 109Japan, 28Jauaperí River, 47-50Jhai Computer, 83-85Jhai Foundation, 83-85Kabul, 93-95Karnataka State, 79-82Kenya, 55, 56, 69-71Khumbu Region, 43-45Kosovo, 89-91land line, xiv, xv, 25, 26Laos, 83-85latency, xiii, 12Latin America, 110Least Developing Countries, 4,

56licensed, 7-14, 17-22Linking Everest, 43-45Linux, 64, 84local loop, xiv, 22, 25, 58, 73, 89Lukla, 43-45Maasai Education Discovery,

69-71Maasailand, 69-71MAC, 8, 10

Malawi, 53-56Manaus, 49MAP, 79-82marginal market, xvi, 39, 57Marubeni, 28McCormack Graduate School,

69-73Media Lab Asia, 25Media Lab, MIT, xiv, 25, 26, 76,

79mesh network, xiii, 11, 12, 108,

109Mexico, 49, 57, 58, 61Microsoft, 21, 22, 90Millennium Development Goals,

99Mobile Access Point, 79-82monopoly, xiv, 17, 18, 32, 34,

57, 58, 63, 64, 66Morocco, 56Mozambique, 63-66Mt. Everest, 43-45multi-hop, xiii, 11, 12National Institute of Commun-

ications, 63-66National Telecommunicationsand Information Administration,

101, 103NATO, 89New York, v, 27, 99NGOs, xi, 1, 34, 38, 47-50, 55,

63, 69, 89Nigeria, 19, 56, 79n-Logue Communications, 73-

76Nongovernmental organizations,

xi, 1, 34, 38, 47-50, 55, 63, 69,89

North America, xiii, 4, 19Norway, 28Novica, 1NTIA, 101, 103off-the-shelf, xv, 37, 38, 83, 100OnSat Network Communica-

tions, 47, 49Organization of American

States, 59packet, xiv, 10, 11, 25, 26, 89,

109Partners in Education, 69-71PDA, 7, 9, 11, 32, 108Philippines, 28Phon Kham Village, 85PHY, 8-10, 13post office, 38, 79, 93, 94Pristina, 89-91radio, v, xiii, 7-14, 18, 20, 21, 26,

64, 83, 100, 101, 103, 107, 108regulatory

activities, 100; authority, xiv-xvi, 17, 19, 20, 29, 39, 64, 99;environment, xvi, 34, 37, 58,61, 89, 90, 94, 101, 107;framework, xiv, xvi, 17, 20, 65;impediments, 61; permissions,29; response, 32; restrictions,61, 71repeater, xiv, 25, 64, 65rural areas, xvi, 1, 11, 22, 27,28, 34, 39, 55, 56, 59, 63, 69,79, 100

Sagarmatha Pollution ControlCommittee, 43SARI, 73-76SatMex, 49Senegal, 56, 69-71Serb, 89Sherpa, 43-45signal-to-noise ratio, 8Solar Electric Light Fund, 47-50South Africa, 19, 55, 56, 63-66spectrum, v, xiii, xiv, 7, 10, 13,

14, 17-22, 26, 29, 33, 34, 48,100, 101

Sprint, 31, 58Stanford Medical Center, 58Sustainable Access in Rural

India, 73-76Sweden, 85Tamil Nadu State, 73-76Tamil Nadu Veterinary SciencesUniversity, 76

1.Pages1_End 11/19/03 10:18 AM Page 126

INDEX 127

telekisoks, xvi, 32, 73-76, 79-82,93-95, 102

telemedicine, 44, 47, 49Telenor, 28telephone, xiii, xiv, xv, 3, 25, 27-

29, 53, 54, 57, 80, 83Telmex, 57, 58train station, 8, 100Tunisia, 56Uganda, 91ultra wideband, 13UMass Boston, 69-71UN Development Program, 93-

95UN ICT Task Force, v, vii, viii,

ix, xi, xv, 99UN ICT Task Force, xi, xv, 99Unicode, 84UniNet Communications, 63-66United Kingdom, 28United Nations, v, 89, 99

United States, 1, 18, 20, 22, 25,28, 53, 56-58, 85, 101

university, 38, 50, 63, 69-71, 73,76, 84, 89, 102

unlicensed, 7-14, 17-22urban areas, 3, 10, 21, 27, 28,

33, 34, 55, 60, 63, 89USAID, 65, 69, 70value chain, xvi, 1, 79, 107, 108,

110Venezuela, 19village, xv, 1, 25-28, 39, 44, 50,

55, 73-85Virtual Yeti team, 43-45voice-over IP, xiii, xiv, 11, 19,

39, 43, 44, 55, 58, 83, 84, 100Voxiva, 91VSAT, 44, 47, 79webcam, 58Wi-Fi Alliance, 100Wi-Fi, v, vii, viii, xiii, xiv, 7, 21,

22, 25, 29, 31, 32, 33, 50, 60,61, 66, 79, 85, 89, 90, 100

WiFinder, 110WiMax Forum, 100WiMax, xiii, xiv, 9, 100, 109WINX, 66Wireless Internet Exchange, 66Wireless Internet Institute, v, vii,

viii, xi, xv, 17, 37, 99, 116WISPs, xv, 17, 18, 31, 37, 53-

66, 101WLAN, 7, 9, 11, 18, 21, 84, 93, 94World Bank, 1, 59, 81World Radio Conference, xiii,

xiv, 8, 21, 100, 103, 109World Summit on InformationSociety, xi, 37WorldLink Communications,

43-45Xixuaú-Xipariná Ecological

Reserve, 47-50

1.Pages1_End 11/19/03 10:18 AM Page 127

1.Pages1_End 11/19/03 10:18 AM Page 128

The Wireless Internet Opportunity for Developing Countries

“Wireless technologies have a key role to play everywhere, but especially in developing countries and coun-

tries with economies in transition. With considerable speed and without enormous investments, Wi-Fi can

facilitate access to knowledge and information, for example by making use of unlicensed radio spectrum to

deliver cheap and fast Internet access. Indeed, it is precisely in places where no infrastructure exists that

Wi-Fi can be particularly effective, helping countries to leapfrog generations of telecommunications tech-

nology and infrastructure and empower their people.”

—From UN Secretary-General Kofi Annan’s welcoming message to

“The Wireless Internet Opportunity for Developing Countries” conference

at UN Headquarters in New York on June 26, 2003

“Wireless Internet may be a very effective and inexpensive connectivity tool, but it does not carry any

magic in itself. It can only be successfully deployed as demand for connectivity and bandwidth emerges

in support of relevant applications for the populations served. These may be supporting e-government, e-

education, e-health or e-business or agriculture applications. But those are not easily implemented in the

developing world.

“One common characteristic of the case studies in this volume is their unconventional, often grassroots ori-

gin. Entrepreneurs from the private, public, or not-for-profit sectors have independently developed original

business models pointing to potential solutions for the developing world. Most, however, must confront seri-

ous challenges that are nontechnical in nature and associated with legacy regulations, administrative obsta-

cles, and the opposition of incumbent telecommunications operators.”

—From the preface to “The Wireless Internet Opportunity

for Developing Countries”

cover 11/19/03 9:45 AM Page 2