Hotspots - Connect the dots for a Wireless Future by klutzfu59

VIEWS: 0 PAGES: 102

									Hotspots – Connect the dots for a Wireless Future
Final Report for Ericsson Business Innovation and Telia Research

Royal Institute of Technology Stanford University 3G Alternatives Communication Systems Design May 2001

David Alvén Resmi Arjunanpillai Reza Farhang Sachin Kansal Nauman Khan Ulrika Leufvén

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

EXECUTIVE SUMMARY
Wireless Internet promises to be the next disruptive technology, leading to big changes in the lives of the end-users and tremendous economic opportunities for infrastructure providers, operators, terminal producers and content generators. The goals of Ericsson and Telia – they want to gain a major share of this emerging market by providing hardware and services that will delight the end user.

The deployment of third generation networks is being questioned more frequently. Strong forces, both commercial and technological, are pushing for other alternatives. Since we started working on this project in January 2001, we’ve noticed a definite winner in the Wireless Internet Space – Wireless local area networks (WLANs). There are several competing WLAN technologies such as IEEE 802.11b, HomeRF and Bluetooth. At this point, IEEE 802.11b is the strongest alternative, with a clear upgrade path to IEEE 802.11a. Today, WLANs can provide data connectivity at up to 11 Mb/sec per access point (IEEE 802.11b); and within 1 to 3 years, they will provide access speeds of up to 54 Mb/sec (IEEE 802.11a and HiperLAN/2) and looking beyond 3 years, this data rate is expected to reach 100 Mb/sec. In contrast to WAP, WLAN does not require any new content creation or application development to attract users. Everyone is looking for the killer wireless application – well, the killer app might very well be the access!

3G will happen; it will be deployed despite the success of alternative technologies. Whether or not WLANs are a threat to 3G is a multisided, complex question. There are a number of possible outcomes to the forthcoming WLAN market.

WLAN coverage can be provided by: 1. One or many fixed broadband ISPs 2. One or many wireless carriers 3. One or many entirely new WLAN providers → Or a combined ISP, wireless carrier, and WLAN provider might control the entire data communications market

2

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

WLAN access can be: 1. Adopted by Wireless Carriers and incorporated as a complementary part of 3G networks → Wireless carriers could then attract more customers 2. Positioned by a WLAN provider as a better alternative to 3G → There would be a battle over wireless customers Seamless handover between: 1. GPRS networks and WLAN hot spots → Might be sufficient for users and prevent an upgrade to 3G → Might prevent 3G operators from charging for services 2. UMTS networks and WLAN hot spots → Might attract users to upgrade to 3G → Might lead to WLAN eating into potential revenues from UMTS In this report, we have come up with two business models based on what we believe is the most feasible outcome. We call the two models Share Point and SwedenOpen.

1.1

Share Point

The Share Point business model is based on the fundamental idea of letting people, businesses and public locations share wireless LAN access with each other and charge each other for this service. The solution makes a lot of sense for unlicensed frequency standards like IEEE 802.11b, as anyone can set up an access point and broadcast on the frequency. Because of these obvious benefits, people are already sharing access today without any support from operators or service providers. However, for usage to really become pervasive, this sharing must be brought about in a structured and organized manner, coordinated by a centralized entity. We have judged the most well positioned primary executor for this business model would be an existing broadband ISP and all the access points put together hence will be called the Share Point Network.

The primary launching pad to grow the user base for this service will be the home and private offices. These customers will be provided wireless LAN hardware at cost prices to encourage adoption. After they are used to this convenience, it is expected that they will naturally want this access in other places. The various segments and their motivations are:

3

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

!

Private Share Zones – Areas with access points in homes, offices etc. Access points are mainly set up for personal use at a favorable price, but when sharing occurs with neighbors, the monthly cost for the access point declines for the owner.

!

Public Share Zones – Areas with access points set up by Share Point in densely populated areas to promote the service. These locations are not owned by profit making businesses. Areas can include parking lots, public parks, bus stations and other gathering points.

!

Commercial Share Zones – Areas with access points paid by commercial businesses like cafés, restaurants, hotels, and stores. The payment share zone could be a profit making investment for businesses in addition to driving traffic to their site. They will earn money on the traffic generated through the access point at their location.

Payment can be done in two ways, either through a service subscription with a monthly bill, or by purchasing a surf card. The surf cards will be sold in retail stores, just like mobile telephone cards are today. Since equipment is being sold at cost, we expect the bulk of the profit to come from these subscriptions and surf cards, as growth will decrease per user costs.

1.2

SwedenOpen

SwedenOpen (SO), a private independent entity will set up and maintain a network consisting of access points all over Sweden, mainly in densely populated areas. These access points intend to cover public places such as Museums, Parks; commercial establishments such as hotels, cafes, and restaurants; and also schools and colleges. All access points will be connected to the SwedenOpen Internet eXchange (SO IX). The SO IX in turn will be connected to the networks of various participating ISPs, who will bring their cable to the IX. The basic ideas of SwedenOpen are: • •

Provide an operator neutral network so that the end-user gets the freedom of choice to pick any operator she wants. Generate revenue through three different types of users: o o People who already have access to broadband Internet at home People who sign up to be members of a particular ISP through SwedenOpen on a monthly basis o People who buy pre-paid Surf Cards to access the Internet via SwedenOpen

4

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

• •

Revenue is shared between SwedenOpen, participating ISPs and the location owners of covered hotspots The first target market will be the City of Stockholm with other cities of Sweden being covered starting Year 2.

1.3

Other Recommendations

Besides executing on the two business models mentioned above, Ericsson and Telia should undertake several other initiatives: • • • • • • • Renting out laptops/PDAs to users at airports, shopping malls, cafeterias, etc. for a short term use Roaming agreements with other wireless access providers Develop seamless handover capabilities between WAN and LAN solutions Be early to market with IEEE 802.11a compliant hardware Bundle hardware with services to provide savings to end users and increase sales Develop capabilities for delivering wireless Internet access to people traveling in trains, cars and airplanes Explore international markets especially rest of Europe and Asia as potential targets for wireless LAN solutions

Broadband Across Transparent LAN

5

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

EXECUTIVE SUMMARY ........................................................................................................... 2 1.1 1.2 1.3 2 3 4 5 6 SHARE POINT..................................................................................................................... 3 SWEDENOPEN ................................................................................................................... 4 OTHER RECOMMENDATIONS............................................................................................. 5

CONTEXT .............................................................................................................................. 9 INTENDED AUDIENCE .................................................................................................... 10 INTRODUCTION................................................................................................................ 11 GOAL OF THE PROJECT ................................................................................................ 14 APPROACH AND METHODOLOGY ............................................................................. 15 6.1 6.2 6.3 6.4 SECONDARY RESEARCH .................................................................................................. 15 PRIMARY RESEARCH ....................................................................................................... 16 CONCLUSION ................................................................................................................... 16 LIMITATIONS TO THIS STUDY .......................................................................................... 17

7

TECHNOLOGY OVERVIEW ........................................................................................... 18 7.1 7.2 7.3 INTRODUCTION................................................................................................................ 18 EXISTING WIRELESS LANS SYSTEMS ............................................................................. 18 “NEXT GENERATION” WIRELESS LANS SYSTEMS .......................................................... 19

8

CURRENT SOLUTIONS FOR WIRELESS DATA ........................................................ 21 8.1 8.2 8.3 WAN WIRELESS DATA SOLUTIONS................................................................................ 21 LAN WIRELESS DATA SOLUTIONS ................................................................................. 22 OUR CONCLUSION ........................................................................................................... 24

9

PLAYERS IN THE WIRELESS LAN MARKET ............................................................ 25 9.1 9.2 9.3 9.4 9.5 CHIP MANUFACTURERS .................................................................................................. 25 HARDWARE VENDORS .................................................................................................... 26 WIRELESS ACCESS PROVIDERS ....................................................................................... 26 ACCESS IN CORPORATE OFFICES .................................................................................... 29 OTHER INTERESTING PLAYERS ....................................................................................... 29 BUSINESS MODELS ...................................................................................................... 32 TECHNOLOGY USED IN BUSINESS PLANS ..................................................................... 32 TARGET MARKET LOCATION ...................................................................................... 33 CONDITIONS FOR SUCCESS .......................................................................................... 33

10 10.1 10.2 10.3

6

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

11 11.1 11.2 11.3 11.4 11.5 11.6 11.7 12 12.1 12.2 12.3 12.4 12.5 12.6 12.7 13 13.1 13.2 13.3 14 15 15.1 15.2 15.3 16 17 18 18.1 18.2 18.3 19 19.1 19.2 19.3

SHARE POINT................................................................................................................. 36 THE SHARE POINT NETWORK ...................................................................................... 37 ROLES AND VALUE CHAIN ANALYSIS......................................................................... 42 FINANCIAL DATA ........................................................................................................ 45 QUANTITATIVE ANALYSIS........................................................................................... 45 TECHNOLOGICAL REQUIREMENTS AND ROADMAP ..................................................... 47 RISKS ........................................................................................................................... 48 ALTERNATIVE/COMPLEMENTARY SHARE POINT IMPLEMENTATIONS ........................ 50 SWEDENOPEN................................................................................................................ 54 SWEDENOPEN NETWORK – A BRIEF DESCRIPTION ...................................................... 54 END-USER EXPERIENCE ............................................................................................... 55 ROLES AND VALUE CHAINS ........................................................................................ 57 ROADMAP .................................................................................................................... 66 SUSTAINABLE ADVANTAGES OF SWEDENOPEN.......................................................... 68 RISKS AND CHALLENGES ............................................................................................ 68 FINANCIAL ANALYSIS .................................................................................................. 71 OTHER IDEAS AND RECOMMENDATIONS ........................................................... 72 ‘ADD-ON’ IDEAS .......................................................................................................... 72 POTENTIAL BUSINESS MODELS ................................................................................... 76 GENERAL RECOMMENDATION .................................................................................... 77 FURTHER RESEARCH ................................................................................................. 78 SOME PARTING THOUGHTS ..................................................................................... 79 WERE THE GOALS OF THE PROJECT REALISTIC? .......................................................... 79 IF WE WERE TO RE - DO THIS PROJECT, WE WILL ….................................................... 79 PROBLEMS FACED ALONG THE WAY ............................................................................ 80 CONCLUSION ................................................................................................................. 82 APPENDIX - ACRONYMS AND ABBREVIATIONS................................................. 83 APPENDIX - SERVICE TRIALS................................................................................... 86 SWEDEN ....................................................................................................................... 86 USA............................................................................................................................. 87 CONCLUSIONS ............................................................................................................. 89 APPENDIX - UNIVERSITY PROJECTS ..................................................................... 90 STANFORD UNIVERSITY .............................................................................................. 90 UNIVERSITY OF CALIFORNIA – BERKELEY ................................................................. 93 MIT ............................................................................................................................. 93 7

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

19.4 19.5 20

CAMBRIDGE UNIVERSITY ............................................................................................ 94 KTH............................................................................................................................. 94 APPENDIX - MATRICES............................................................................................... 98

20.1 EXHIBIT 1: WIRELESS LANS AND PAN STANDARDS COMPARISON MATRIX .............. 98 20.2 ............................................................................................................................................ 98 20.3 EXHIBIT 2: MAJOR WIRELESS WANS STANDARDS COMPARISON MATRIX ................. 98 21 22 23 24 24.1 24.2 24.3 24.4 24.5 24.6 BIBLIOGRAPHY............................................................................................................. 99 WEB RESOURCES ......................................................................................................... 99 ACKNOWLEDGEMENTS ........................................................................................... 100 PERSONAL THOUGHTS OF THE TEAM MEMBERS .......................................... 101 DAVID ALVÉN ........................................................................................................... 101 RESMI ARJUNANPILLAI ............................................................................................. 101 REZA FARHANG ......................................................................................................... 101 SACHIN KANSAL ........................................................................................................ 101 NAUMAN KHAN ......................................................................................................... 101 ULRIKA LEUFVÉN ...................................................................................................... 102

8

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

2 CONTEXT
This report is the product of collaboration between two groups of students and faculty from the Royal Institute of Technology (KTH), Stockholm, Sweden, and Stanford University, USA as part of a course “Communication Systems Design – 2G1319”. This particular project is co-sponsored by Ericsson Business Innovation and Telia Research. The purpose of the course was to enable the participants to solve a “real world” problem, demonstrate independent learning skills and effective project management; show communication skills while presenting the solution and work as a successful team.1 This particular team, unlike most of the other teams in the course, was spread geographically over two continents and hence got an opportunity to learn about crosscultural and geographical factors in teamwork. The core team consisted of three students from KTH and three from Stanford, working over a period of five months.

1

From the course website ‘http://www.online.kth.se/courses/2g1319/indexe_en.html’

9

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

3 INTENDED AUDIENCE
The intended audience of this report is the executive teams of Ericsson Business Innovation and Telia Research led by Mr. Mats Segerström – Investment Manager at Ericsson Business Innovation, Mr. Jasminko Mulahusic, Research Engineer at Telia Research and Mr. Roland Larsson, Business Development Manager at Telia. Ericsson Business Innovation can use this report to identify potential investments in the wireless LAN market. They can spin out companies from within Ericsson Business Innovation based on the suggested business models. Ericsson can also leverage the useful information in this report to market their wireless LAN hardware in the most relevant market segments. Telia can get a better understanding of the evolving role of a traditional operator with the advent of new technologies such as wireless LAN, and use the recommendations and suggested business models to maintain their leadership in the fixed and mobile data markets. They can increase customer penetration through compelling services at an appropriate price. The report will enable both Ericsson and Telia to recognize potential synergies amongst the two companies, and further strengthen their supplier-customer relationship. Readers of this report will gain: ! An understanding of the wireless data market in general, and the wireless LAN value chain in particular ! An understanding of the major end-user needs and habits that drive wireless Internet usage ! An insight into the opinions and concerns held by the leaders in this industry ! An understanding of potential profitable business models and other ideas and recommendations, on which they can initiate action ! Ideas about what future research to carry out in the wireless data markets ! An understanding of how geographically-separated project teams work effectively

10

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

4 INTRODUCTION
A number of trends can be detected in the world of communications. First of all, the number of different devices a person uses to communicate is increasing rapidly. A couple of years ago it was uncommon for people to have more than a PC. Today, people use a handful of devices, from stationary PCs and laptops to PDAs and mobile phones. Not only have the number of devices increased, but they have also decreased in size. More and more devices are small enough to carry with us, i.e. more of them are becoming mobile, and more and more of them are used by only one user, i.e. more of them are becoming personal. These changes affect the way we use communication devices and it creates the need for new applications. Another ongoing trend is the diminishing gap between telecommunication and data communication. Most people today agree that telecommunication is moving closer to data communication rather than the other way around. Voice, that used to be the predominant payload in our communication networks, is being replaced by multimedia. Voice only takes up a fraction of the bandwidth that intense data applications take up, so when the bandwidth increases, voice will only take a small percentage of this. Another trend of convergence that we can see is between fixed and wireless networks. There will always be some physical facts that differentiate air from other fixed transport media, which we cannot do anything about. But there are a number of ways to work around these flaws, and engineers worldwide are extremely busy coming up with new solutions. In order to increase bandwidth, cells are becoming smaller and smaller. We’re moving from macrocells, to microcells, to picocells, most probably to femtocells2. New algorithms are being implemented to overcome the high air transmission error frequency. A number of new wireless technologies are present or emerging: ! GSM will become GPRS and/or EDGE, and will soon be further enhanced with UMTS (3G) in some places. ! Wireless LAN (WLAN) solutions include IEEE 802.11b and IEEE 802.11a, HiperLAN/2, etc. ! Low power radios like Bluetooth and HomeRF. ! Cellular Digital Packet Data (CDPD), multihop (Ricochet) networks. ! DECT (cordless) and IrDa (Infrared). All these trends are changing the structure of the market place. There is a possibility that companies based on today’s technologies might lose substantial market share if they cannot foresee or at least keep up with these trends. The bright side is that there is a vast amount of money to be made for the company or companies who do master this difficult task.
http://www.online.kth.se/courses/common/2g1319/pdf/mobile-lecture.pdf, Mobile Personal Computing and Communication Lecture notes of G. Q. Maguire Jr.
2

11

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

We believe that within a couple of years, we will have ubiquitous access via a wide array of devices, ranging from phones and PDAs to watches and house equipment. There are however a number of questions that needs to be answered in order to predict a successful business model based on today’s trends and future visions. ! First of all, timing is critical. You must see the window of opportunity and catch it. Is now the time for ubiquitous access or will access in some specific areas be sufficient, and in that case which areas? Is the window of opportunity two years from now? What business model can be successful built on today’s technology? While considering time, the chain of services must also be evaluated. There is no reason for high bandwidth if there are no applications or devices that can take advantage of it. One example of this is Europolitan’s GPRS roll out. They are losing money because the majority of mobile phone manufacturers haven’t marketed their GPRS phones yet. Closely related to this are the benefits of being first to market. A company might have to make sacrifices in the initial phase due to low usage in order to ensure being first on the market. There are a number of questions regarding the cost – How much will it cost? Who will be willing to pay for this and how? What are the profits? What is the volume? Can in be a stand-alone business, or will it just be an add-on to existing services? What roles do different companies play in this model? Who delivers the services? Will it be carriers, telephone companies, the streets department, contracting companies, power companies, coffee shops, or even clothes manufacturers? Who will own the user data and user interaction? What are the presumptions of culture? Will people accept the system? If implemented, what would the impacts of the culture be? The cultural aspect is closely related to time and fashion. It is critical to know how long it will take for people to adapt. This can be clearly seen with mobile phones; they have become an important part of our image. Where is the market? Who will the users be? Another issue related to people acceptance is education. Will the system be easy enough for users? What resources will be needed to use, install and maintain the system? An ironic example of this today is mobile phone address books. There are a surprisingly large number of people who put stickers with handwritten phone numbers on the phone, instead of entering them into the phone’s address book. Last, but maybe the most important matter of them all, is technology. Technology is the basic foundation for the model. Advances in technology can either be forced by a need, 12

!

!

!

!

!

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

or they can be achieved because the technology is possible. In the latter case, the hope is that the technology will drive the need. Some people claim that there is no need for widespread network access; that the technology is possible, but that there is no need for it. This might be true at the moment, but on the other hand, the need might not emerge until the service is there. Take the Xerox machine as an example. There was no need for copying paper because no one had ever though of it. We don’t think anyone questions the need for the ability to copy paper today. These are some of the questions that we will try to answer in this report.

13

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

5 GOAL OF THE PROJECT
The objective of our project is to develop one or more feasible business models for alternative solutions to 3G (Third generation mobile) networks. The business models might build on a combination of various technologies such as UMTS, Wireless LAN, fiber backbones, satellite etc., or the result might be based on a single, superior technology. Like any other business model, we attempt to address the following points in our model(s): ! ! ! ! ! ! Value proposition for the players in the market i.e. infrastructure providers, mobile operators, service providers, content providers and of course, end users Design of different value chains Role of established and new players Revenue model for each player Technological implications of our proposed model Risks and barriers to our plan

14

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

6 APPROACH AND METHODOLOGY
In order to achieve the goals of the project that as listed above, our approach was to use several methods and several sources of information and opinion. Our research methodology has been to first acquire all background information about the technologies that can be considered an alternative to 3G. Then we took a snap shot of the current market to see what companies are working on practical applications of these technologies. After identifying these companies, we set up interviews with representatives from both their marketing and their engineering departments. In order to live through a first hand end user experience, we conducted service trials with various different service providers both in Sweden and the US. For better understanding of the end users’ needs, we conducted an end user survey, both in the US and in Sweden. In this section, we shall describe our secondary and primary research methodology. In the end, we discuss the limitations of this study.

6.1

Secondary research

We used several methods to the necessary background information for this project:

6.1.1 Web Research
We used the World Wide Web extensively to find information about various fields of interest. The Web helped us get up to speed with the differences between various existing technologies such as IEEE 802.11b and 802.11a, Bluetooth, HiperLAN/2 and others. We also researched the different players in the market to study their business models. Ongoing research at various universities has been studied to see if and how it is relevant to our project.

6.1.2

Industry Reports

We have gathered market research that has already been done in this field to avoid re-inventing the wheel (See Section 18).

6.1.3 University Projects
Since we study at two of the most prominent universities for engineering and business, it was relatively easy to set up interviews with people involved in various relevant projects at these universities, as well as to acquire information about already conducted projects. We looked at projects at both KTH and Stanford, as well as other universities. The location of these two universities, in the Silicon Valley and Stockholm/Kista put us right in the middle of the research center for mobile Internet services in the world (See Section 22).

15

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

6.1.4 Matrices
Based on our research, we constructed two matrices, one describing the players in the market, and the other describing the different technologies and their current status in the market. The use of matrices facilitates the comparison between different players and technologies (see Section 23).

6.2

Primary research

In addition to relying on the already existing research, we also conducted our own research at the grass-roots level:

6.2.1 Interviews
We conducted several interviews with representatives from various companies, and technology experts at universities. Being students at prominent universities and interesting subject matter were two factors that eased the process of setting up interviews. Most of the companies we talked to were most forthcoming and helpful (see Section 24).

6.2.2

End user Survey

We conducted a web based end user survey with the purpose to find out what the end users want, which ultimately, is the driver of all businesses. We had two different versions of the same survey - one for the American market and the other for the Swedish market. We also conducted a “on the road” version of the survey during which we talked with various people in both the US and Sweden, and had them fill out the surveys in person. Those results were then entered into the online versions. We got over 500 responses to the two surveys combined (see Section 25).

6.2.3 Service Trials
We tried several different existing wireless LAN services both in the Bay area in San Francisco and in Stockholm. This helped us assess the end user experience and get a first hand feel of the problems an end user faces while using WLAN (See Section 21).

6.2.4

Brainstorming Sessions

Through out the course of the project, we had regular conferences via telephone and video. During these conferences, we have had long brainstorming sessions. We also spent a week together in March when the Stanford part of the team visited Sweden during which time we had several very productive brainstorming sessions. These brainstorming sessions were very helpful in coming up with a thorough end user survey, to present ideas to each other, come up with feasible business models.

6.3

Conclusion

After the secondary and primary research, we gathered sufficient background knowledge and understanding of the WLAN market to help us come up with feasible business models. This analysis and the technical implications will be further discussed under Section 10. 16

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

With the help of all the above-mentioned activities, we analyzed the different business models that we generated. The question that we asked ourselves was - “Is this model technically feasible and is the model viable from a business point of view?” This process led us to our two final business models, the Share Point model (see Section 11) and the SwedenOpen model (see Section 12).

6.3.1

Technical Implications

After having generated the business models, we analyzed the technical implications these models would have. “Will it work today? Or do we need to invent something in order to make it work?” “Will the AAA issues be solved?” “When will roaming between GPRS and WLAN be solved?” “Will the end user be willing to pay for the hardware or does it need to be subsidized?” “How much interference will there be if IEEE 802.11b gets too popular?” The technical implications will be discussed further in the business models.

6.4
!

Limitations to this study
Narrow geographical focus. This is due to the fact that we had limited time and resources; and because Sweden and the USA are the current major innovation centers for WLAN services. Apart from these two countries, we also looked at some solutions in Finland. Not focused on security. Security of WLANs remains a big issue that needs to be solved. Though we have recognized the problem, and suggested a few possible solutions, we have not attempted in this project to solve the security issues. The results of the end-user survey are angled more towards students and professionals between 20-30 years of age rather than a true representation of the overall population. It is also biased towards technology savvy people. We only considered viable and currently existing technologies for our business models. Some technologies such as Bluetooth, IEEE 802.11a and HiperLAN/2 are not available for mass deployment yet. No number on infrastructure costs. We do not have exact costs of hardware when bought in bulk, though we have tried to “best guess.”

!

!

!

!

17

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

7 TECHNOLOGY OVERVIEW
We believe that for a business model to succeed, an investigation into existing technologies and relevant players is necessary. In this section, we present a brief overview of all the relevant technologies in the market. Please refer to Section 17 for definitions of several terms used and explanations of acronyms.

7.1

Introduction

Wireless LANs were first introduced in 1997. Initially, there were two different technologies enabling wireless LANs, one for corporate environments (IEEE 802.11) and the other for home networks (HomeRF). Different companies supported one of the two technologies. Today there are three major Wireless LAN technologies: IEEE 802.11b, HomeRF and Bluetooth. We should clarify that Bluetooth has been recently elevated to the ranks of being a WLAN technology, even though it started out as just a cable replacement technology and it was designed to offer point-to-point links. The ability of Bluetooth to support WLAN environments still remains to be proven. There is an ongoing discussion in the Bluetooth SIG (Special Interest Group) to support wireless LAN applications in the next generation Bluetooth technology. The HomeRF consortium has started to develop HomeRF/2, which will be run on the same frequency as HomeRF and would support a data transfer rate of 10Mbps. In September 1999, IEEE approved a revision of the IEEE 802.11 standard, called 802.11b or 802.11 “High Rate” that provides much higher data rates (5.5 and 11 Mbps), while maintaining the 802.11 protocols. At this time, several companies started to adopt this new “fast” technology for on-campus wireless networking. As time passes, more and more users want to have wireless access at different locations such as home, hotels, airports, etc. We believe that IEEE 802.11b was able to beat out HomeRF by the virtue of being the first on the market with a fast access of 11Mbps.

7.2

Existing Wireless LANs systems

Below is a brief introduction to some of the already existing technologies:

7.2.1

HomeRF

As the name suggests, HomeRF was developed from the beginning to bring wireless networking to the consumer in his home using RF (Radio Frequency). HomeRF products operate in the globally available 2.4 GHz ISM (Industrial, Scientific and Medical) band using FHSS (Frequency Hopping Spread Spectrum). First generation HomeRF products have peak data rates of 1.6 Mbps and cover virtually all homes and small offices with a 50-meter typical indoors range. Second generation HomeRF products (which were supposed to be on the market by mid-2001) use 10 18

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Mbps peak data rates while still providing whole home coverage. Third generation HomeRF devices are planned to be even faster (20 Mbps) and are supposed to be on the market in the second half of 2002. HomeRF is fully backward compatible. Cayman Systems, Compaq, Intel, and Proxim are some companies that work with HomeRF, though Intel recently announced strong support for IEEE 802.11b.

7.2.2

IEEE 802.11b

In September 1999, IEEE approved 802.11b to create a standards-based technology that could span multiple physical encoding types. This approval added two higher speeds, 5.5 and 11Mbps, to 802.11. The 802.11b standard is designed to have a transmission range of about 30 to 100 meters (300 feet) and operate in the 2.4-GHz ISM band using DSSS (Direct Sequence Spread Spectrum Technology). The standard uses a CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) and as Ethernet, 802.11b uses an identical MAC (Media Access Control). Designers also included a shared-key encryption mechanism, called WEP (wired equivalent privacy), in the specification. The WEP mechanism covers station-to-station transmission. The standard specifies usage of the RC4 security algorithm. Efforts are underway to boost up the performance of 802.11b standard to speeds of 22 Mbps or even up to 54 Mbps, and this new protocol will be called 802.11g. Lucent Technologies, Intersil Corp, Cisco and Symbol are some of the major companies that support the 802.11b standard.

7.2.3

Bluetooth

Bluetooth is a low cost and low power wireless connection method with a small footprint that makes it very well suited for cable replacement. The idea that resulted in Bluetooth was to make a wireless PAN (Personal Area Network) with a transmission range up to 10 meters. Bluetooth was born in 1994 at Ericsson mobile communication. In February 1998 five companies, Nokia, Ericsson, IBM, Toshiba and Intel, formed the Bluetooth SIG (Special Interest Group). Bluetooth communication occurs in the unlicensed ISM band at 2.4GHz. The transceiver utilizes frequency hopping to reduce interference and fading. The communication channel can support both data (asynchronous) and voice (synchronous) communications with a total bandwidth of 720 kbps.

7.3

“Next Generation” wireless LANs systems

Further research is being carried out for a better and faster wireless LAN system. Here are two of the major solutions slated for release in the not too distant future: 19

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

7.3.1

IEEE 802.11a

IEEE ratified the 802.11a at the same time as the 802.11b standard in 1999; its goal was to create a standards-based technology that could span multiple physical encoding types. IEEE 802.11a is designed to have a transmission range of 30 up to100 meters, to support 54Mbps. The IEEE 802.11a standard operates in 5-GHz UNII (Unlicensed National Information Infrastructure) band, which also is free for the end users. Like IEEE 802.11b, 802.11a uses MAC (Media Access Control). However, IEEE 802.11a uses an entirely different encoding scheme, called OFDM (Orthogonal Frequency-Division Multiplexing), which departs from the traditional spreadspectrum technology. The OFDM scheme was intended to be friendlier to office environments. Both security and QoS will be better in IEEE 802.11a based LANs. Atheros, Radiata (recently acquired by Cisco), Lucent and Cisco are some of the companies that support this technology.

7.3.2

HiperLAN/2

HiperLAN/2 (HIgh PErformance Radio Local Area Network type 2) is an ETSI (European Telecommunications Standards Institute) project called BRAN (Broadband Radio Access Networks), developing a new generation of standards, which will support both asynchronous data and time critical services (e.g. packetized voice and video) that are bounded by specific time delays to achieve an acceptable QoS. The HiperLAN/2 Global Forum was launched in September 1999 and was supported by six founding members: Bosch, Dell, Ericsson, Nokia, Telia and Texas Instruments. HiperLAN/2 provides a flexible platform for a variety of businesses and home multimedia applications that uses the unlicensed 5GHz UNII. It also supports a set of bit rates up to 54 Mbps and a transmission range of 30 up to 100 meters. To achieve 54Mbps, HiperLAN/2 makes use of a modularization method called OFDM. This network will support both authentication and encryption. For a comparison of various WAN and LAN systems see Section 23.

20

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

8 CURRENT SOLUTIONS FOR WIRELESS DATA
In our quest for an alternative to 3G, we tried to look at several existing solutions for wireless data delivery.3 Wireless networks can be broadly classified into Wide Area Networks (WAN) and Local Area Networks (LAN).

8.1

WAN Wireless Data Solutions

There are several ways a user can access data services over a wide area network:

8.1.1 Internet on your wireless phone…. or is it??
Cellular wireless networks today use Wireless Application Protocol (WAP) and the Short Messaging Service (SMS) channel of GSM, CDMA, TDMA networks to deliver mobile data. WAP is a mobile data technology which makes graphic-less Web content available on WAP enabled cellular phones. Even though WAP enabled cell phones can potentially access a large number of websites, the interface is not very user-friendly. Also, the transfer speed is 9.6 kbps – 14.4 kbps, which is frustratingly slow. Wireless operators have not been able to provide intuitive and attractive applications to their users, making WAP very unpopular. SMS is a text messaging service available on the European GSM networks. The service is very popular with billions of messages being sent and received every month. Even though this service has received tremendous user response, it does not provide access to the Internet. NTT Docomo’s i-Mode service in Japan demonstrates a successful implementation of a data service on mobile phones. Within a few years, the i-Mode service has signed up 23 million users.4 Even though i-Mode users do not have access to the whole Internet, there are about 42,000 iMode sites available that provide relevant and targeted content to users generating high levels of usage.

8.1.2 Use your Wireless Phone as a modem
Many users access the Internet by hooking up their laptop computer or PDA with their mobile phone that acts a wireless modem. Using a cable, or the infrared ports of these devices can make the connection. The main problem with this solution is that the speeds are very low. Also, people do not like to buy and carry extra cables for connection, nor is the infrared connection very reliable requiring continuous line of sight. Bluetooth promises to solve these connection problems, but no one knows when Bluetooth will arrive in the market.

3

Taking into the account the audience of this report, we feel it is unnecessary to discuss these solutions in depth. In this section, we mention several wireless data solutions with a brief analysis. 4 Hirose, Norihiko, VP, Business Development and Investment at NTT Docomo USA and founding member of I-Mode team – in a presentation at the “Wireless Internet” symposium at Stanford University on May 16, 2001.

21

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

8.1.3

Palm VII and Wireless Modems for PDAs

Palm VII provides instant connectivity as soon as the user raises the antenna of the device. The service is called Palm.Net and is available in several major metropolitan areas in the US. However, it only provides access to a limited number of partner sites and not to the whole Internet. Also, the form factor of Palm VII is too big, and does not make it a very popular device for people who want to put their PDA in their pocket.5 Companies like Novatel and Sierra Wireless sell wireless modems (like the Minstrel modem by Novatel) for PDAs – both Palm OS and Pocket PC based devices. Companies like Omnisky and GoAmerica provide Service. Again, these services provide only selected content that has been customized to fit the PDA form factor. These services have not been very popular among the users because of the bulky form factor of the wireless modems and the cost (both service and modem.)6

8.1.4 Metricom Ricochet
Metricom provides a wireless access service called Ricochet, which allows the user to browse the Internet at speeds of about 128kpbs. Metricom’s Micro Cellular Data Network is the only wireless network designed for data from the ground up. The user has to buy a Ricochet modem or a Ricochet wireless card, which can then be used with a laptop computer or a PDA by paying a flat monthly fee of about $75. The service is available in several metropolitan areas in the US and the coverage is pretty uniform in these areas. Even though users like this service, Metricom has not been able to sign up enough subscribers. Since building a network from ground up is a capital-intensive job, and considering the fact that Metricom is running low on cash, there is a good chance that the company might not survive till the end of 2001.7

8.1.5

Email pagers

Blackberry, by Canada based Research in Motion, is a palm-sized pager that allows the user to send and receive emails in an always-on mode. The device is quite popular among business people in the US. However, the keyboard is very small and it allows only email communication and a few other selected services.

8.2

LAN Wireless Data Solutions

Lately, there has been a lot of activity in this category. The most common way to access a wireless local area network is by inserting a network interface card in the laptop (or PDA) which
5

Because of excess inventory in May 2001, Palm VII is being sold at a “throw-away” price of $99 down from $330 at one point in time. 6 In May 2001, Omnisky started a rebate scheme in which they are giving away a $299 value modem for free if the user signs up for a one-year service plan. This scheme hints at the lukewarm user response that the wireless modems for PDAs have received. 7 Please note that this is a speculative statement being made on the basis of conversation with a former partner of Metricom (identity not disclosed).

22

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

then communicates with a base station (Access Point) located in the vicinity. The most common protocols used for this are IEEE 802.11b and Home RF, which have been discussed in the previous section on Technology Overview. The user gets a fast broadband (11mbps) access to the complete Internet but coverage area is small – if the user moves 100 feet away from the access point, he is certain to lose the connection. There are several settings in which this solution is being used:

8.2.1

Public Hotspots

Wireless access providers are setting up access points in places like airports, train stations, hotels and cafeterias allowing their subscribers to browse the Internet at these locations. Most of the providers such as Telia HomeRun in Sweden, and MobileStar and Wayport in the US are targeting the business traveler, as they are likely to be the early adopters for this service. Rapid proliferation of these hotspots is currently underway. Currently, the problem with this solution is that they are narrowly targeted towards the business travelers and hence other segments are not fully aware of its existence. Also, it is limited to the people who carry laptops and are willing to buy a wireless card (in some cases, the wireless cards are provided for free by corporations to their mobile employees). Some of these solutions are very expensive. A case in point is Telia HomeRun, which charges approximately $150 per month for unlimited usage at any of their hotspots.

8.2.2

Offices and Universities

Several small, medium and large sized corporations are installing wireless LAN access points within their premises, thus enabling their employees to access the Internet and the corporate database even while being away from their desk. They also provide free wireless cards to the employees. Some corporations buy mass subscriptions from providers like MobileStar and give them to their employees who can then access the Internet from hotspots covered by the provider. This is the market that most of the access providers are trying to target. Several universities such as KTH in Stockholm are putting up access points on campus and are providing free broadband wireless access to their students, faculty and staff.

8.2.3 Free networks
Several user groups such as Electrosmog in Stockholm and the Bay Area Wireless Users Group (BAWUG) in the San Francisco Bay Area are installing access points at several locations with the aim of letting anyone use the networks for free. They generally set up access points in houses or on rooftops. Though the mission is very noble, we feel that this is not a scalable solution and these self-run networks will never be able to provide reliable service to people.8
8

This point has been validated by Brewster Kahle, founder of SF LAN, a wireless user group in San Francisco, CA in a telephone conversation with us.

23

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Some people with wired broadband connections in their home are putting up access points for their own in-house use – this allows them to move around the house and even outside the house for a certain distance and stay connected all the time. We feel that in-house wireless access is a big potential market and we have developed a business model named ”Share Point” aimed at capturing this market (see Section 21).

8.3

Our conclusion

In summary, we can say that there are two types of wireless data solutions out there today: ! ! Wide area solutions that are slow speed, and sometimes give access to only limited content. Local area solutions that are high speed and give access to the entire Internet but are limited in range.

People desire high-speed access to the complete Internet everywhere they go. We believe that rapid and wide proliferation of wireless LANs will provide the desired solution. Even though people will not get ubiquitous access, the areas of coverage (number of hotspots) should be increased to cover most of the places where people converge. All service providers are targeting the business traveler market today – we feel that the general user should be targeted too. We have taken this approach in coming up with our business models, which are discussed later.

24

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

9 PLAYERS IN THE WIRELESS LAN MARKET
In the previous section, we mentioned the various wireless data solutions currently present in the market. In this section, we will focus on the players that form a part of the wireless LAN value chain. Figure 1 illustrates the wireless LAN value chain, as it exists today (the direction of the arrows depicts the flow of money through the value chain). As of today, public wireless LAN solutions are being provided in offices and public hotspots such as airports, hotels, etc.

9.1

Chip Manufacturers

As the name suggests, the chip manufacturers produce the chips that form the heart of the wireless LAN hardware equipment. AMD and Intel are examples of companies involved in the manufacture of IEEE 802.11b compliant chipsets, while companies like Atheros, Radiata (recently acquired by Cisco) and Resonext design and manufacture IEEE 802.11a chipsets.

Chip Manufacturers

Infrastructure and hardware vendors

Hardware Distributors or Resellers

Public Hotspots

Access Providers

Corporate offices

End user

Figure 1 Current Wireless LAN value chain

25

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

9.2

Hardware Vendors

Next in the value chain come the hardware vendors that supply base stations, network interface cards (commonly just called ‘wireless cards’) and other components such as bridges and repeaters. This market is very crowded with at least 25 vendors selling IEEE 802.11b compliant hardware components9. Some of the most prominent players in this market are Cisco, Symbol, Nokia, Ericsson, Lucent, Proxim, etc. Efforts are being made to embed IEEE 802.11b chipsets directly into portable computers – Dell Computers is expected to roll out IEEE 802.11b compliant portable computers by the end of 2001.10 In May 2001, Handspring announced an IEEE 802.11b compliant Springboard Module for their Visor line of Personal Digital Assistants (PDAs).11

9.3

Wireless Access Providers

With the emergence of wireless LAN technology, a new breed of service providers is being created whom we call the wireless access providers. These wireless access providers specialize in different segments of the hotspot market such as airports, or the hospitality industry. Their business model is simple – buy the hardware, install the hardware at the hotspots, sign up users and charge the users for wireless Internet access. From our interviews with several access providers, we have found that the agreements between these providers and the hotspots are not standardized and they vary on a case-by-case basis. Two of the most popular types of agreements are: ! ! The hotspot owner bears the cost of installing the hardware and the monthly broadband access charges, and shares the revenues generated by Internet usage at his premises. The hotspot owner does not bear any cost, but does not get any share of the revenue. He may or may not get any upfront fee from the access providers.

We will briefly describe some of the companies in this category12:

9.3.1

Telia HomeRun

Telia HomeRun is a part of Telia Mobile and is focused on providing secure wireless Internet access to corporate customers. Their hotspots include airports in Sweden, SAS lounges worldwide, hotels and convention centres. They currently have around 160 sites covered with their wireless Internet access solution and aim to have at least 400 sites by the end of 2001. They

9

IEEE 802.11a components are not available in the market yet, and are expected to be rolled out by the end of 2001. Source: Presentation by Theresa Meng, CTO, Atheros Communications, at the “Wireless Internet” symposium at Stanford University on May 16, 2001. 10 IDC Research “Unwiring the Network: Worldwide Wireless LAN Market Forecast Update, 2000-2005”, April 2001. 11 Source: http//www.handspring.com 12 This is not an exhaustive list of companies in this space. We are describing the companies that we interviewed personally or on the phone. ‘Access providers’ are the most interesting node of the value chain for us and hence we go into more detail.

26

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

offer different pricing schemes, ranging from an initial fee of 2495 kr and monthly fee of 1495 kr down to 24hr surf card for 120 kr. They are located in Nacka, outside Stockholm, Sweden.

9.3.2 PowerNet
Their business idea is to offer wireless and mobile broadband services. They offer wireless solutions for indoor use, outdoor use and public surf zones. They also offer fixed Internet access, city networks as well as wireless corporate networks. PowerNet provides fixed broadband access wirelessly by using omni directional antennas to transmit the signal to their customers’ directed antennas. They plan to put into operation various applications based on wireless broadband communication in all the Nordic countries by the end of 2001. The home package costs 2000 kr to sign up and then 295 kronor / month for 24 months which includes all the hardware needed. They have different pricing schemes when selling to corporate houses. They are located in Stockholm, Sweden.

9.3.3 WirelessBolaget
They offer wireless solutions for indoor use, outdoor use (point-to-point, point to multipoint etc.) and public access zones (Nokia’s PAZ). WB’s primary focus is to provide wireless access to businessmen. At the moment they are installing IEEE 802.11b WLANs in hotels. WB uses Telenordia as the Internet Service Provider. They outsource installation, maintenance, support and training of hotel personnel. They offer different prices depending on whether a use has a WLAN card or not –195 kr package includes WLAN card, user name and password enabling unlimited surf for 24 hours. (They have recently introduced additional offers for 2,3 and 4 days). There is a 95 kr package that includes user name and password enabling unlimited surf for 24 hours, but does not include renting the WLAN card. They are located in Hammarby, outside Stockholm, Sweden.

9.3.4 MobileStar
MobileStar is the one of the biggest wireless access providers in the United States. Their hotspots cover common hotel areas (lobbies, restaurants, meeting rooms, etc.), airport gate and terminal areas, airline lounges, coffeehouses and restaurants. They are targeting the mobile professional by primarily selling directly into big corporations, which enables them to sell several accounts in a single sales process. In January 2001, MobileStar announced a strategic deal with Starbucks and Microsoft that will entail providing broadband access in every Starbucks location in the US. They also have a roaming agreement with SkyNet Global, a wireless access provider operating in the Asia-Pacific region. They have several pricing schemes ranging from $2.50 per 15 minutes of use, to $59.99 for unlimited usage per month. MobileStar is based in Texas, United States.

27

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

9.3.5 Wayport
The target user group and locations for Wayport are similar to the ones for MobileStar. Wayport and MobileStar and striving to outdo each other in terms of location capture and user subscription. Wayport already has presence is some hotels in Europe and Canada, through a strategic deal with the Four Seasons Hotel chain. Recently, Wayport won the rights to provide wireless access at the San Jose and San Francisco International airports, which was a major victory. They charge a pre-paid fee of $49.95 for 10 airport connections or 6 hotel connections. They are also based in Texas, United States.

9.3.6

Surf and Sip

Surf and Sip is a small regional player based in San Francisco, California, United States. They are primarily targeting ”high-loiter” locations such as cafeterias and some restaurants. Their target user group is students and other people who like to sit and work at one place for hours at a time. Though they only have a regional footprint in some of the cafes in the San Francisco Bay area, they are expanding their reach into the other US states. They are offering free access to anyone who signs up for the service by May 31.13

9.3.7 Airwave
Airwave is also a small regional player based in Palo Alto, California, United States. Their target locations are cafes, restaurants, bookstores and university campuses in the San Francisco Bay Area. Until early May 2001, Airwave charged $10 per month of unlimited usage with the first month free. However, in mid-May, they announced that they were scaling back their operations by reducing the number of covered locations, and made the network completely free from everyone. We infer that this is a very capital-intensive business and several companies are finding it difficult to raise additional capital during the current negative market sentiment.

9.3.8 Softnet (Aerzone)
Aerzone, a subsidiary of Softnet Systems was focused on providing broadband wireless access at airports, hotels, conventions centres, etc. During late 2000, Aerzone struck strategic deals with Delta and United Airlines, and also won the rights to cover several US airports such as San Francisco and Denver. However, in December 2000, Softnet decided to discontinue the wireless operations of Aerzone because of unexpected amount of capital requirements and failure to raise new money. However, Aerzone has a division called Laptop Lane, which provides virtual offices for business travellers at several airports in the United States with high-speed Internet connections, fax and teleconference facilities.

13

They extended this deadline twice from March 31 and then April 31. We think that they have been unable to create enough visibility due to resource constraint, resulting is low subscription levels. At each of their locations, they also provide an Internet kiosk – basically a wirelessly connected laptop computer that customers can use for a fee.

28

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

9.4

Access in Corporate Offices

Several small, medium and large sized corporations are installing wireless LAN access points within their premises, thus enabling their employees to access the Internet and the corporate database even while being away from their desk. They also provide free wireless cards to the employees. Some corporations buy mass subscriptions from providers like MobileStar and give them to their employees who can then access the Internet from hotspots covered by the provider. This is the market that most of the access providers are trying to target.

9.5

Other Interesting Players

In addition to the players mentioned above, there are several other interesting players that are relevant to our research. We describe them briefly:

9.5.1

A Brand New World

A Brand New World was founded in 1994 under the corporate name Radio Design TJ AB as a research and development enterprise, active within mobile telephony systems (NMT 450). The current operations, focused on mobile Internet systems and infrastructure, have been pursued since the beginning of 2000. ABNW launched a terminal with what they call 4G technology during first quarter of 2001. The 4G technology is based on GSM and WLAN (IEEE 802.11b). Basically, they have integrated a GSM phone and IEEE 802.11b PCI card into a PDA. They see themselves as an alternative and complement to 3G.

9.5.2

Sydkraft

Sydkraft is a Swedish power company that is starting to offer broadband to their customers through the power outlets. Every household is connected to a net station where the voltage is reduced from 10 kV to 230 V. One net station serves between 2-300 households. Sydkrafts intention is to bring a broadband fiber to the net station and from there give the customers access through the outlets. Since all households are connected to a net station, Sydkraft has hereby solved the “last mile” problem. The special type of technical equipment that is required hasn’t been available until now. The end-user will get a data speed between 1 and 2 Mbps. Costs of the required hardware will be included in the initial subscription fee. The monthly cost hasn’t been decided yet, but it will be a flat rate.

29

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

9.5.3

Jippii Group

They are a Finland-based multi-services operator that has launched WLAN services for both the consumer and the corporate market. They are pushing WLAN as an alternative to both 3G and fixed Internet access. The company aims to become a full multi-services data provider in both the fixed and mobile sectors. They have presently launched three city networks in Finland and are testing 48 other cities. Both the consumer and corporate customers pay US$35/month. The company aims to be one of the top five ISPs in Europe. The technical investments are as important as network rollout. Although a soft handover between access points belonging to the same network is part of the IEEE 802.11b standard, users may not be able to roam over networks belonging to different operators. Jippii’s roaming technology allows users to switch between any available WLAN access points, regardless of who operates that particular network. “We’re looking to roll out our technology anywhere, even if that means other networks making the money,” says COO Matti Roto14. The roaming technology is available either under license or as a complete package.

9.5.4 Metricom (Ricochet)
We have already described this company in the previous section called ”Current Solutions”.

9.5.5 ArrayComm
ArrayComm produces software technology that uses multiple antennae base stations to improve performance in congested radio cells. This reduces dropout of calls, mitigates co-channel interference from nearby cells and even allows for reuse of channels within cells. Over 75,000 of these base stations have been deployed worldwide. ArrayComm's new technology i-Burst is an IP-centric wireless architecture designed to affordably extend the broadband Internet experience into wireless. ArrayComm claims that iBurst enables the deployment of low cost, ubiquitous, high speed wireless data networks (2 Mbp/s) with massive capacity at price/performance levels equal to or better than those of today's wire line broadband alternatives. These claims will be tested in the first test deployment in San Diego later this year. It is important to keep an eye out on i-Burst, as could be either strong competition or natural upgrade for our Share Point model in the future.

14

Mobile Internet Magazine, Volume 2 Number 7, April 6 2001. p 9-10

30

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

9.5.6 HereUare
HereUare (pronounced Here You Are) provides a suite of services that can aggregate the wireless networks of several providers and provide global roaming capabilities to their customers. They provide customization capabilities that enable service providers to implement special privileges, protection, and restrictions for segmenting their users. In addition, they provide private label billing capabilities for wireless service providers. They enable location owners to control the initial “splash” page that appears upon access, providing them with a marketing tool for acquiring and retaining customers. HereUare is based in San Jose, California, United States.

31

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

10 BUSINESS MODELS
The fundamental goal of this study is to come up with self-sustaining profit or non-profit models that can provide broadband wireless access to a large portion of the population with some portability, if not mobility. After several months of research of wireless technologies, companies and trends, we have come up with two business ideas that we believe can popularize and help maintain a sustainable user base of wireless data users. We call the two models Share Point and SwedenOpen. Share Point can be implemented in a number of ways with different stakeholders and value chains. We will give a brief overview of these options and discuss when and how we believe the various implementations will be successful. The first two sections describe things that are common for both the business models.

10.1 Technology used in business plans
We strongly believe that Wireless LANs will become a definite complement to 3G, if not an alternative. Both our business models use Wireless LANs to provide wireless Internet access in numerous areas. Today, wireless local area networks (LANs) can provide data connectivity at up to 11 Mb/sec per access point; within 1 to 3 years they will provide access speeds of up to 54 Mb/sec and looking beyond 3 years this data rate is expected to reach 100 Mb/sec15. The technology used in the initial phase will be the IEEE 802.11b or Wi-Fi standard. The 11 Mbps data rate provided by IEEE 802.11b is significantly more than the 2Mbps promised by 3G. There are various practical reasons that make IEEE 802.11b the clear choice in today’s market for popularizing wireless access technologies. IEEE 802.11b hardware is cheap, it provides high bandwidth, uses an unlicensed frequency band, is easy to deploy and has by far the most momentum in terms of adoption. In fact, we expect the current market activity to be the beginning of a meteoric rise for the standard. According to analyst firm Cahners In-Stat Group, the market for the technology is expected to reach $2 billion in 2002 and $4.6 billion by 200516. Of course, IEEE 802.11b is not the best technical solution as nothing ever is, and we have detailed its various drawbacks in (see section Error! Reference source not found., Error! Reference source not found.). Our business models take this view into account and make room for the adoptions of, and migration to newer, better technologies like HiperLAN/2 and IEEE 802.11a, as they are developed and commercially introduced. The value propositions in our business models are never inextricably tied to the technology and this makes the solutions

15

Victor Bahl, Anand Balachandran, Srinivasan Venkatachary. The CHOICE Network: Broadband Wireless Internet Access In Public Places. February 2000 Technical Report MSR-TR-2000-21 16 Cisco details wireless strategy, http://cnet.com/news/0-1004-200-5964269.html?tag=nbs, May 17, 2001

32

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

adaptable to unanticipated changes. Nonetheless, we have sufficient confidence in the IEEE 802.11b technology and it is unlikely that there will be a need to replace it very soon.17 There has recently been an increasing concern amongst access providers on developing “killer applications” that would get the users hooked, and bring in a steady revenue stream from them. This is valid thinking for networks like the wired Internet, which are in more mature stages. However, we feel that wireless data access is in such a nascent stage that access itself is the killer application. After people are enabled with enough bandwidth, they can decide the most appropriate uses for it, themselves.

10.2 Target Market Location
The business models are oriented towards Sweden and Scandinavia. This has less to do with why other countries are not suitable and more to do with the conclusion that Sweden is ready for such a revolution. The mass adoption of wireless phones and sophisticated Internet access on wired networks makes it ideally positioned for a wireless Internet revolution. The hype around 3G and its delay has made people even more anxious for wireless access and signs of the WLAN fever catching on can already be witnessed in the market. We are confident our business models can easily be adapted to other countries for replication and further expansion; however this will not be covered in the current report.

10.3 Conditions for Success
There are a number of factors that will contribute to our business models’ success. Our business models primarily focus on the wireless infrastructure, not wireless devices. Devices are needed in order for the infrastructure to be widely used. We have listed some accelerators and inhibitors in the following paragraphs that we feel are relevant to our models.

10.3.1 Accelerators
+ Prices on WLAN cards are falling and hardware manufacturers are starting to integrate WLAN chips directly into computers, PDA modules18, Compact Flash cards19, etc. We believe, that within a short period of time, prices on WLAN cards will fall to an acceptable price range for “normal” users.

As protocol wars on the Internet between TCP/IP and ATM have proven, arguably better standards can sometimes fall by the way side when put head to head with the most widely adopted competing standards. There is a good chance the new challengers to 802.11b might lose based simply on their delayed time to market.(Death to ATM, http://www.redherring.com/index.asp?layout=story&channel=70000007&doc_id=1000016100 ) 18 Intel Transforms Handspring Visors Into Wireless Communication Tools; SpringPort Wireless Ethernet Module Enables 802.11b Connectivity for Handspring Visor Handhelds May 8 2001. http://www.zdii.com/industry_list.asp?mode=news&doc_id=BW20010508BW0434&pic=Y&ticker= 19 Symbol Introduces Compact Flash Wireless LAN Adapter To Connect PDAs To 802.11b Wireless Networks May 8, 2001. http://www.symbol.com/news/pressreleases/pr_retail_compact_flash.html

17

33

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

+ +

Wireless LANs are addictive. Once you’ve tried it in one place you want it in more places. Cisco conducted a study on Wireless LAN usage that involved 20 businesses with 1,000 or more employees that had been using WLANs for at least one month20. The results were significant. Time saved by WLAN use ranged from one to 15 hours per user, per week, with an average time savings of eight hours per user, per week, according to survey results. WLAN is strongly supported by Microsoft. WLAN support is being built into future Windows and PocketPC operating systems for an easier user experience (see section Error! Reference source not found., Error! Reference source not found.).

+

+ + + + +

Sales in the rest of the world (ROW) are enhancing growth opportunities for WLAN, moving away from its once U.S.-centric, vertical model. Mobile travelers, accessing the Internet or enterprise virtual private networks (VPNs) from public places, will fuel sales of WLAN solutions. Higher-speed (IEEE 802.11a) products to hit the market will rectify sluggish data rates for WLAN. Leading vendors will aggressively invest in WLAN through company acquisitions, research and development, and industry alliances. WLAN shipments are experiencing record growth in traditional and nontraditional WLAN markets. IDC forecasts even stronger growth once IEEE 802.11a and HiperLAN/2 products reach the market21. WLAN hotspots in airports, hotels, train stations etc., will fuel the growth of WLANs in other areas. Unlike WAP (Wireless Application Protocol), which has failed till now to prove mass adoption, the high bandwidth that WLANs provide will be sufficient for surfing regular Internet sites with an HTML browser. This lowers the barrier for end user acceptance, since no new knowledge is required.

+ +

10.3.2 Inhibitors
− − Semiconductor shortages for mobile applications and telephony could slow future growth of WLAN. The inception of Bluetooth products in 2001 could cause havoc in WLAN development through interference in the 2.4GHz band.

"Wireless LANs: Improving Productivity and Quality of Life, 05/17/01 http://www.commweb.com/article/COM20010517S0003 for more news on the study see http://newsroom.cisco.com. 21 The majority of accelerators and inhibitors are from the report Worldwide Wireless LAN Market Forecast and Analysis, 1999-2004, IDC Report #23431 - December 2000.

20

34

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

−

Many standards are present today, confusing the average consumer (as discussed in section 7, Technology Overview). Without standardization, many potential WLAN users will be lost in the shuffle. The 5GHz products have not reached production levels of the 2.4GHz products. Crowding of the 2.4GHz band, without a current 5GHz solution, is strangling future development of WLAN. An increased focus on marketing, rather than research, could disrupt current technological development of WLAN. Future development of the industry will hinge on regulatory bodies such as the FCC and others. Strained relations between WECA, HomeRF, and Bluetooth will only hurt the industry. Greater cooperation is needed between the organizations.

−

− − −

35

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

11 SHARE POINT
The Share Point business model is based on the fundamental idea of letting people share wireless LAN access with each other. The solution makes a lot of sense when you consider the characteristics of WLANs. Because IEEE 802.11b uses the 2.4GHz unlicensed ISM band, anyone can set up an access point and broadcast on the frequency. However, there is a limit to usage in one location, as interference will occur if the frequency band is cluttered. Unlike wired networks, wireless technologies are perfect for bandwidth sharing since they are pervasive. In addition to this, IEEE 802.11b does not require line of sight and the radio waves at that frequency can penetrate walls with considerable success. Because of these obvious benefits, people are already sharing access today without any support from operators or service providers. For usage to really become popular however, this sharing must be brought about in a structured and organized manner, coordinated by a centralized entity. This will not only help resolve the related economic issues in a professional and impartial manner but will also effectively mitigate the technical risks arising from shared usage. Most recently, the immense value of wireless LAN access points has become obvious to a growing population of users. This ‘enlightened’ group has started various efforts to put up access points either for themselves or to share with others on a somewhat “philanthropic” basis. A host of start up businesses has also jumped into the foray, focusing mostly on providing wireless access in public or commercial locations. As yet they have not been able to scale their individual business models, largely because of a lack of sufficient revenue and large upfront hardware costs. The Share Point idea capitalizes on early end users and others from a broad cross-section of society who is adequately prepared to adopt something of this nature. Share Point will provide them with a platform to share their local networks as part of a much larger conglomeration that increases access opportunities for all. Combining all the WLANs into an organized effort provides users with fast Internet coverage in many areas. We call this collection of organized WLANs the Share Point Network. The Share Point network will provide various advantages beyond the obvious: 1. Increased adoption of Internet and location related services while mobile, due to ease of use (in contrast to WAP which puts a higher demand on users) and increased bandwidth. 2. Increased Internet availability for those who don’t have it at home or want to use it sparingly at a lower cost. 3. Extra incentive for people to upgrade from dial in modems to broadband cable. 4. Opening up a new market for location based services and applications. 5. WLANs have a short range and this will force deployment focus in areas that would have the most potential in terms of usage. Hence, creating an efficient paradigm for utilization of expensive hardware resources.

36

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

We believe that networks in homes and especially wireless networks will grow dramatically over the next few years (see Section Confidential). Today WLANs are increasingly being installed in offices; soon employees will want to have it in their home. When users have it in their home, they will become used to it and demand to have it in other locations. According to a study by Network magazine22 connecting roaming users and providing WLAN access to conference rooms where the two top choices for WLAN installation in companies. However, 65% of the companies asked would not open up their own WLAN for non-employees. The key to encourage access point installation is to offer every one in the value chain a good deal. This implies offering end users subsidized WLAN hardware, discounted subscriptions, sharing revenue with access point location providers and partners etc. If the offers are attractive enough, the Share Point Network will be built by a number of parties, thereby increasing the deployment speed and coverage of the network. What these offers are, how they are paid for and whom the players will be is described below.

11.1 The Share Point Network
The Share Point business model is designed with an existing broadband ISP in mind as the primary executor. Since the business model is based on sharing data access, the fixed broadband provider is critical to this model for economic, legal and logistical reasons. We chose to apply this model directly to Telia ISP for the sake of relevance and because Telia has the biggest share of the Swedish ISP market, 31% in December 200023. While we will illustrate with suggestions on how a proprietary advantage can be maintained in this space, the same business model could just as easily be picked up and practiced by another Swedish or international ISP. The Share Point Network will consist of three different zones, which will have different roles in the value chain and the cost and revenue sharing will be different for each situation. Subscription packages will treat each of these zones separately for granting access. ! Private Share Zones – Areas with access points in homes, offices etc. Access points are mainly set up for personal use at a favorable price, but when sharing occurs, the cost of the access point declines for the owner. Because the fixed access infrastructure to homes varies significantly (see Table 4: Internet customers per access. Source: Post- och Telestyrelsen.) We will use access points with multiple ports so that the fixed broadband connection can be either Cable-TV, ADSL, ISDN or Ethernet LAN. Public Share Zones – Areas with access points set up by Share Point in densely populated areas to promote the service. These locations are not owned by profit making

!

22 23

The Pulse. Networking Magazine, November 2000. Den svenska telemarknaden fortsätter att växa. Post och Telestyrelsen. March 9, 2001.

37

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

businesses. Areas can include parking lots, public parks, bus stations and other gathering points. ! Commercial Share Zones – Areas with access points paid by commercial businesses like gas stations, cafés, restaurants, hotels, and stores. They will earn money on the traffic generated through the access point provided at their location. The Payment Share Zone will be a profit making investment for businesses, in addition to this, they will provide a value add to regular customers.

The payment and settlement models for the above areas (to be explained later) have been designed to cater to their respective needs. The picture on the next page should better help illustrate the different categories.

38

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Parking Lots
Building 1 Building 1
REST AREA

Bus stops

House

Rest Areas
House

Parks

Private Share Zones

Public Share Zones

Commercial Share Zones

Gas station

Motel

Stadium
Outdoor mall

Figure 2: Share Point Network Surf Zones

39

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

11.1.1 Pricing and Payment methods
We’ve based the prices on what we think are reasonable premiums to be charged for wireless Internet access without jeopardizing mass-market acceptance. Payment can be done in two ways, either through a service subscription with a monthly bill, or by purchasing a surf card. The surf cards will be sold in retail stores, just like mobile telephone cards are today. The surf card will have a username and a password that will be valid for the number of minutes displayed on the cards. The user is charged per 15 minutes, so you can use a 3-hour surf card maximum 12 times. The surf cards are critical for Share Points’ success. Mobile telephone cards have become a huge success; they stand for 44% of the mobile telephone subscriptions in Sweden24. Surf cards will be sold with different values, for a variety of surf hours. The surf card gives access to the entire Share Point Network i.e. Private, Public and Commercial Share Zones. For those who use the Share Point Network often and for those who feel it is easier to be charged monthly, a monthly subscription will be available as well. The customer can pay for any combination of the following package of hardware and services. Under almost all circumstances, aggregated monthly charges should not ever go above the psychological price point of 500 kronor.

24

Den svenska telemarknaden fortsätter att växa. Post och Telestyrelsen. May 9, 2001.

40

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Offers to Households
Type of service Fixed broadband Characteristics Charges A fixed broadband connection provided Confidential by Telia ISP to the home. A fixed broadband connection provided by Telia ISP to the home and an installed WLAN access. Provides Confidential unlimited surf at home for all family members. (Deduction if neighbors sign on, see below).

Private Share Zone

A fixed broadband connection provided Access to Neighbor’s by Telia ISP to the home. Unlimited Confidential wireless surf using neighbor’s access Private Share Zone point. Add on to any of the above A WLAN card (for stationary, laptop or Confidential PDA). Table 1: Share Point offers to households

WLAN card

Offers to end users
Type of service Characteristics Charges Access to the entire Share Point Network (Private, Public and Commercial Confidential Network Share Zones). 10 hours limit for each increment of subscription and higher overuse charges above that. The ability to purchase a prepaid card for a nonShare Point Surf Card subscriber. Provides access Confidential to the entire Share Point Network (Private, Public and Commercial Share Zones). Table 2: Share Point offers to end users

Share Point Access

41

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

The subscription for the Private Share Zone owner will be decreased for every neighbor that signs an Access to Neighbor’s Private Share Zone subscription. This achieves both hardware cost efficiencies but will also subsidize home purchase of equipment if access point owners are compensated proportional to the number of neighbors sharing their access point. This is most important because it links customer satisfaction with rewards for the access point owner who might otherwise have no incentive to share access or ensure a good user experience for his/her neighbor. In this case, if the neighbor is dissatisfied with the wireless access point, he will unsubscribe and hence cause a monetary loss to the access point owner with the Private Share Zone.

Offers to become a Commercial Share Zone
Type of Service Fixed broadband Description Cost Revenue A fixed broadband Confidential connection provided by Telia ISP to the location.

Telia lends out an access Commercial Share point to the location at no Confidential Zone charge. Revenue Sharing Telia will share the revenue with the Commercial Share Zone. Confidential

Table 3: Share Point offers to commercial establishments

Share Point will give the Commercial Share Zone X kronor for every hour a user surfs. Even through the revenue is incremental per user, it will add up as the service becomes popularized. A café that has 15 users per day, that surf 30 minutes each, will make X kr/month. In order to attract locations to become Commercial Share Zones, Share Point will lend the location an access point and install it at no charge. A criterion for being a Commercial Share Zone is that you generate at least 15 hours of total usage every month, starting two months after the installation. Assuming the location meets the criterion; all the location provider will have to do is sign a 24-month agreement to become a part of the Share Point Network.

11.2 Roles and Value Chain Analysis
The following picture gives an overview of the involved players, and how the money flows between them. The players are numbered and described in detail below the picture.

42

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

H. G.
Private Share Zone (Household with access point.) Neighbor without access point

Commercial Share Zone

F.

Telia AB

A.

Hardware +$ Web Content Provider

D.

Telia ISP

B.

Share Point

C.

Surf Card Resellers (Pressbyrån etc.)

WLAN Maintenance

E.

Hardware Public Share Zones Hardware vendor

WLAN Installation

I.

Figure 3: Share Point Value Chain

A. Telia AB: Telia ISP and Share Point are two business units within the Telia AB enterprise. Users will never know about the business units. They will know about the “Share Point Network” which is something that Telia has. B. Telia ISP: Telia will interface households and end users and offer the new Share Point services. Telia will take care of billing through their existing billing system. Telia ISP will charge customers for fixed broadband connection as well as any additional Share Point services. These will be paid through a monthly or quarterly invoice to households and end users. C. Share Point: The Share Point unit handles the wireless Share Point Network. This includes operation of the user authentication system, supervision of installed access points (mapping), revenue sharing, surf card distribution, hardware manufacturer agreements, content creation, access point installation and maintenance. Installation, maintenance and possibly even web content creation should be outsourced. 43

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

D. End User: End users can either buy a surf card from a surf card reseller or a Commercial Share Zone, or subscribe to the service through Telia. E. Surf Card Reseller: The surf cards should be sold in every store that sells mobile telephone cards today as well as all the Commercial Share Zones. The surf card reseller buys the card from Share Point and sells them at a 20 % margin. The margins depend on the location and availability of the reseller. F. Commercial Share Zones: Include restaurants, gas stations, movie theatres, shopping malls, sport arenas, etc. Share Point will share revenue with these Commercial Share Zones for traffic passing through access points that they provide at their premises. The Share Point authentication and tracking system keeps track of the total number of minutes that users surf in a particular zone. The user is charged for every 15 minutes of access. The Commercial Share Zone will have to pay for the fixed broadband service. This however can be offset by the fact that the location provider can use the broadband access for internal applications such as inventory and customer care, they can sell ad space on their local web site and provide their customers with online information about products, special offers etc. At the initial phase before the service is widely used, the revenue sharing is deducted from the broadband invoice. When the revenue precedes the fixed cost the profits are transferred to the Commercial Share Zone’s account. G. Private Share Zones: In the Private Share Zones the access point is not lent out but instead sold to the user. The price of the hardware is divided into monthly payments signed on a 36-month contract. WLAN cards can also be offered to end users. Just as in the Commercial Share Zones, anyone who sets up a Private Share Zone will benefit from sharing the access with neighbors. The Private Share Zones are not only for households. Businesses might want to put up access points in conference rooms, consultant offices and show rooms. H. Neighboring Households: Households that want access to a Neighbor’s Private Share Zone at home can sign up for a subscription through Telia. Users can buy surf cards to surf through the neighbor’s share point; however this will not be cost efficient in the long run. I. Public Share Zones: The Public Share Zones are exclusively taken care of by Share Point. The Public Share Zones will be at locations that have a need for access but cannot be made into a Commercial Share Zone. Public Share Zones include locations like parks, parking lots, bus stations etc.

44

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

11.3 Financial Data
The following consumer numbers have provided us with a basic understanding of the ISP subscriber market in Sweden. Internet customers per access 2000-12-31 Private 1 898 400 78 100 16 500 22 100 62 900 470 2 800 76 800 2 158 100 164 600 Business 197 700 107 500 39 000 4 100 0 13 900 6 800 369 000 24 800 Total 2 096 100 185 600 55 500 26 200 62 900 470 16 700 83 600 2 527 100 189 400

PSTN (Modem up to 56 kbps) ISDN GSM xDSL Cable-TV Satellite Other fixed access < 2 Mbps Other fixed access Total number of Internet customers

Thereof customers with broadband * * Cable-TV, xDSL, and other fixed access.

Table 4: Internet customers per access. Source: Post- och Telestyrelsen.

While the number of users currently subscribing to broadband is low, this is anticipated to grow significantly as last mile deployment increases and applications become more bandwidth intensive. Even looking at current broadband adoption rates, we see that the home should be the first target market for broadband wireless due to existing penetration rates. While some may argue that non-portable desktops are more prevalent in the homes and this significantly reduces the need for a WLAN, we feel the argument is a little misguided as most business users who work at specified desks in their office would bring their notebooks home and would want wireless access for ease of use in any part of the home. Finally, price data for existing broadband charges and hardware etc. was obtained from rates quoted in the market and is listed where necessary in our calculations.

11.4 Quantitative Analysis
Customers that sign up for the Share Point Service can choose between different offers as mentioned above. Please refer to section, Confidential for a detailed layout of how the quantitative pricing analysis has been made and justified for these service plans. The following details explain the rationale and some of the key assumptions made in the process. 45

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

The pricing model is cost based because the goal is to popularize the technology and gain a solid foothold in the market. Our quantitative model hence establishes the minimum price that should be charged to customers without making explicit losses on the service or product itself. Hence, the hardware prices are based on COGS and do not include other overhead and operating costs. Having such cost pricing means that there is little subsidy included in the hardware or service prices. There would be some who would advocate going further and taking a loss on the hardware cost in order to gain the fastest possible market foothold. This is unwise, as it should be expected to recoup at least the full cost of the equipment over a 3-year contract maintaining a healthy cash flow. In the current economic environment, it is not advisable to have business models that need unusual infusions of cash in order to scale without generating much revenue in the short term. Hence, we have avoided building a model that is a black hole for cash resources. According to our Swedish survey (see section, Confidential), an overwhelming 66% of respondents showed a willingness to pay 100 kronor or more per month for using wireless Internet services. All of our pricing keeps this in mind. Additionally, there is an overwhelmingly large number of the population in US and Sweden (in the 80% range) who want to pay flat fees as opposed to variable per usage fees. We have tried to accommodate that in most of our pricing as well. However, it is important to address here why unlimited mobile access cannot be provided for a fixed monthly charge unlike all the other services and hardware in Table 1: Share Point offers to households. This is because of our usage based revenue sharing scheme, which gives the right incentives to commercial share point owners to promote the service and gain from its success. If Commercial Share Zones were to be paid a fixed amount every month, they would lose incentive to serve customers well and popularize the service. Also if incremental usage doesn’t pay Commercial Share Zone owners they have no incentive to have someone sit and take up space for example, in a café and result in no revenue. To accommodate both these needs, we have introduced a fixed charge that guarantees service for a certain number of hours and can be bought in various multiples for increased usage. The right question finally is how will the business model eventually become profitable if most pricing is cost based? As the growing numbers of home wireless LAN users get hooked to their wireless access, they along with other mobile users will increase their usage of Commercial and Public Share Zones. The per-user cost of operating this access point network will go down as the number of users increases and we expect the bulk of the profit to come from the resulting Share Point Network access subscriptions and surf cards. We have modeled these variable and fixed costs of deployment under the “Roaming services” heading and have tried to include all costs associated with operating the Share Point Network. As can be seen in year 3, the per user costs fall below most of our roaming price points. Assuming that at least a fraction of the user base will pay for a multiple of the base 10-hour subscription, this should result in fairly high margins and profits for our network and should provide an ample cash flow to finance further growth. 46

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

11.5 Technological Requirements and Roadmap
There are a number of technological barriers that have to be overcome in order for the Share Point model to work. A rather complex tracking and billing system must be implemented; however, we do not by any means feel that this will be unachievable. Changes in the market place, success of the model, as well as technical progress will put high demands on the back-end Share Point system. Not all of the desired functionality has to be implemented at the launch of the service; we will continue to improve the back end system over time. 1. The following functionality is required at launch: # Authentication through username login and password (we strongly believe that users will accept this at the initial phase, because the number of hotspots will be rather limited). # Global user authentication database for scalable solutions. A number of Universities and companies have come up with solutions including Flying Linux25 from KTH, SPINACH from Stanford (see section 19 Appendix University Projects), and CHOICE,26 from Microsoft and the University of California at San Diego. # Surf card minute usage and expiration tracking. # Minute tracking of subscribers in Private, Public and Commercial Share Zones, for monthly charges and revenue sharing. (We can accept a minute tracking fault tolerance of 5 minutes since we charge customers for 15-minute intervals). # Web site that shows areas of Public and Commercial Share Zones. 2. The following features are desired but not critical at launch: # The solution should not be hardware dependent. Any access point should be able to be used in the Share Point Network. This will enable users who already own installed access points to become Private Share Zones and share access with others. This subscription form would have to be added. # SMS service that provides users with information about the nearest Share Point location. # Application that can tell users how much is left on the surf card or how many charged minutes that are added to the monthly bill. 3. As the competition increases and new players enter the market we will have to: # Sign roaming agreements with other WLAN providers

25

Escudero A, Pehrson B, Peletta E, Vatn J.O, Wiatr P. "Wireless access in the Flyinglinux.NET infrastructure: Mobile IPv4 integration in a IEEE 802.11b". 11th IEEE Workshop on Local and Metropolitan Area Networks. LANMAN2001. Co. USA. March 2001. http://www.it.kth.se/~aep/grs/ 26 Victor Bahl, Anand Balachandran, Srinivasan Venkatachary. The CHOICE Network: Broadband Wireless Internet Access In Public Places. February 2000 Technical Report MSR-TR-2000-21

47

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

# Implement authentication architecture that handles user databases operated by multiple vendors. # Improve billing system to 1-minute fault tolerance in order to faster detect change of access point usage. # Quality of Service – guaranteed bandwidth allocation for owners of Private Share Zones and Neighbor subscriptions, “leftover bandwidth ” for surf card customers and other Share Point subscribers. Subscription price could be tied to guarantee bandwidth allocation. 4. As seamless GPRS/UMTS and WLAN handover emerges we will have to: # Implement automatic authentication. Users that are always “online” either through WLAN or GPRS cannot be asked to log on every time the device is used. # Partner with wireless carrier, or distribute Share Point Network subscriptions through wireless carriers. # Combined GPRS/UMTS and WLAN billing solution. A SIM card authentication mechanism for identification and billing could be used. This will be a natural way for the wireless carriers to adopt this business model and it also opens up new fields of applications. The user would be identified and authenticated through the IMSI number in the SIM card. Nokia already has WLAN products with SIM card solutions including WLAN cards with a slot for a SIM card27.

11.6 Risks
There are always a number of risks associated with a business model. Many different factors need to be considered in order to make a viable business model. In this section these risks will be discussed further.

11.6.1 Technological Implications
First and foremost, consumers must be made aware of the limitations of the technology. When introducing a new product, it is better to under promise and over deliver and grows by a positive reputation, rather than disappoints customers and get a negative customer sentiment to prevail in the market. Hence, the following disclosures must be made openly: 1. The bandwidth is not dedicated. Every user shares a total of the fixed broadband connection up to a maximum of 11 Mbps, and effective bit rate goes down exponentially with number of users. 2. Range and bandwidth are calculated under optimum conditions and can vary significantly depending on a variety of factors, including interference and medium of transmission. 3. Interference should be expected and accepted as part of a low cost solution. 4. Security is not guaranteed. Users will have to use VPN connections to secure data.
27

Nokia C110/C111 Wireless LAN Card. http://www.nokia.com/corporate/wlan/card_c110.html

48

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Share Point must also be aware that there are various limitations that the technology places on its own growth and scalability. Not more than 3 access points can be placed in the same area despite no other interference. The total number of people that can share one access point varies drastically in different research reports. Compaq says that 50 high-end users or 150 low-end users can share one access point28. According to our technology expert at KTH, Alberto Escudero, up to 30 users can share one access point. This could be a problem because it means immense success of the service could lead to its own demise! Other scalability and security issues are also still unresolved by the standards bodies or the hardware manufacturing companies.

11.6.2 Financial
The largest financial risk is that customers would either not want to pay for our service at all, or think that it is too highly priced. As we learnt in our interviews with Andrea Goldsmith, Donald Cox and others, this has traditionally been the primary cause for the demise of wireless data systems. However, the failure of these previous attempts at providing wireless data can easily be attributed either to the difficulty/cost of usage or the lack of applications in all the cases that we examined further29. We are confident, based on the results of our extensive research and comprehensive survey, that the market landscape for wireless data services has changed significantly with the wider and pervasive adoption of the Internet. Users have become more and more used to having wired access in the home, office and in other public places, and are now anxious to be able to use the same information and applications when on the move. Both in the United States and in Sweden, the largest number of respondents said they would be willing to incrementally pay $10-$20 for wireless services and we have kept this as a necessary guide to our introductory pricing plans. While we have tried to have conservative estimates for the cost of equipment that we are going to sell, we feel that a ramp in volume will likely drive bulk prices for IEEE 802.11b equipment below what we have estimated in the financial exhibit. Alternatively, a spike in prices of equipment could be caused by a shortage in parts, however this is highly unlikely considering the spare capacity in the industry, standardization of components and the current downturn in related industries that could have competed for the same parts. In the event of such an eventuality, we suggest not changing the price points but altering the access point and wireless card customer agreement, leaving ownership with Telia ISP at the end of the three years so that the salvage value could be applied towards mitigating the higher costs borne for the hardware. Even in and of itself, “leasing” out the equipment rather than selling it to the customer could prove to be a beneficial strategy since this will make customers more receptive to hardware upgrades in the future. If people own the equipment, they will be more reluctant to migrate to new technologies
28 29

Compaq Wireless LAN prestanda. http://www.compaq.se/10produkter/10D0wireless/10D0prestanda.asp Size of WAP screens and the cumbersome nature of data entry on mobile phones was the fundamental reason quoted by interviewed users for not wanting to pay for such existing data services.

49

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

and to pay for new cards. In fact, businesses and self-employed consumers might even see leasing as a plus since leased equipment costs can be counted as expenses for tax accounting purposes. However, eventually most customers might see non-ownership as the lower value proposition that it is and get put off by that. For this reason, we have opted against this strategy unless the cost structure forces us to adopt it. Since there is a current land grab happening for wireless LAN locations in many developed countries, eventually national and international roaming agreements between different service providers will have to become a reality. In fact, roaming agreements between these companies are not only desirable but also necessary for a global solution. However, if the wireless voice carriers are any example to go by, such a system of agreements could take many years to negotiate and execute. This bureaucracy should be kept in mind. Also, we anticipate that this and other integration with the wireless carriers will add significant fixed and overhead costs, however, since such a change is not expected in the next 3 years, it does not figure in our quantitative model. However, it could be a realistic cost to watch out for if the service succeeds to a level where such integration happens.

11.7 Alternative/Complementary Share Point Implementations
We believe that in 2-3 years, access to the Share Point Network will be offered through wireless carriers. This is based on the theory that WLANs will become an integrated part of 2.5 G and/or 3G. As such, the wireless carriers have a natural part in the Share Point model. The idea is based on a fundamental change in the way we access the Internet. As it is now, we pay an ISP for Internet access when we are at home working on our stationary computer or laptop. When we are away from a fixed ISP location we use a GSM modem in our laptop and pay the wireless carrier for access to the Internet. New devices like PDAs, handhelds, web tablets etc., make the line between fixed and wireless Internet very fuzzy. Users should not have to worry if they are using fixed or wireless Internet access. Other reasons for going through wireless carriers, in addition to the conglomeration of networks, is access to their enormous customer base and the advantage of using their established customer support and billing systems. Just to illustrate this customer base, it is important to point out that by the year 2000, mobile subscriptions in Sweden had reached 6 338 000, or 70% of the total population; achieved at a staggering yearly growth rate of 24 percent30. As a result, today, there are more mobile than fixed telephone subscriptions in Sweden.

11.7.1 Share Point Partnership with Telia Mobile
To increase the overall revenue for Telia AB as an enterprise, we recommend that Share Point partner with Telia Mobile. If Telia were to effectively leverage its joint ownership of fixed, WLAN and GSM/GPRS/UMTS networks, it would become the strongest player in all the markets with the ability to offer its customers the entire portfolio of data and voice access. The fact that

30

Den svenska telemarknaden fortsätter att växa. Post och Telestyrelsen. 9 maj 2001. 50

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

fixed ISPs and wireless carriers are partnering or acquiring one another is a validation of this market prediction31 (see Section 19.5). In this model Telia Mobile would handle the Share Point subscription to end-users (Table 2: Share Point offers to end users. Telia ISP would handle the household offers (Table 2: Share Point offers to end users, and Share Point would handle the offers to the commercial establishments (Table 3: Share Point offers to commercial establishments). In this model, the user does not have to make the distinction between a fixed or wireless network. The user simply makes the distinction of being home (fixed) or mobile (away from home). When the user is at home, he/she can surf on PDAs, laptops, anything with a WLAN card in it, on a fixed cost monthly subscription with Telia. As soon as the user leaves the house (or the office where WLANs are paid by the company), access to the Share Point Network is paid to Telia Mobile (could eventually be charged in a single bill). Though the market might not be ready for this yet, GPRS/WLAN devices will soon hit the market. It is essential that Telia have a subscription plan ready for these devices. At the initial phase, we suggest that only one mobile subscription be made available. As WLAN usage becomes more pervasive, different levels of subscriptions could be added for peak hour rates, favorable locations and catered to specific customer segments like business men, young people etc.

Orange and NTL Join Forces. May 14 2001, http://www.thestandard.com/article/0,1902,24449,00.html?mail=1 51

31

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

11.7.2 Value Chain and Roles
If a partnership is done with a wireless carrier, the value chain will look slightly different. Telia ISP will never interface the end user; this will be done by Telia Mobile. The wireless carrier shares the revenue with Share Point.

Surf Card Resellers (Pressbyrån etc.)

E.

Wireless Carrier Telia Mobile and/or others

Commercial Share Zones

F.

H.
Neighbor without access point. Telia ISP

Telia AB

A.

Hardware +$ Web Content Provider

B. G.
Private Share Zone (Household with access point) Hardware Public Share Zone

Share Point

C.

WLAN Maintenance

WLAN Installation Hardware vendor

I.

Figure 4: Share Point Value Chain with wireless carriers.

11.7.3 Share Point Distribution through Multiple Wireless Carriers
If the goal is to maximize the revenues for Share Point specifically, we recommend that Telia use as many wireless carriers as possible as a distributor of Share Point subscriptions and services. The subscriptions and services would include access to the network through an additional charge to the monthly mobile telephone bill. Share Point would charge the wireless carriers a fixed percentage of usage of the Share Point Network, and the wireless carriers could then charge their 52

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

customers whatever they feel appropriate. The profits for Share Point will be less per user in this model, since the revenues will be split with the wireless carriers as well as the Commercial Share Zones. However, the total revenue will be more considering the added number of users.

Comviq

Telia Access Point User with WLAN phone, PDA, laptop etc.

Internet

Europolitan

Telia Mobile

Share Point Server with authentication and mapping of user to wireless carrier.

Figure 5: Architecture for Share Point with Wireless Carriers

53

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

12 SWEDENOPEN
SwedenOpen is an extension of StockholmOpen.net that’s being pioneered by the city of Stockholm and KTH, along with other partners like Stokab, Svenska Bostäder, and hardware vendors and ISPs. We assume the reader is aware of the basic theme behind StockholmOpen.net. (If not, please see www.stockholmopen.net for more background information.) The basic idea is an operator neutral network so that the end-user gets the freedom of choice to pick any operator she wants. Today, a user is forced to use the ISP (wire line or wireless) that provides service in a particular area. For instance, if Telia provide the cable to the City Hall in Stockholm, a person cannot use Bredbandsbolaget when he is there. Our aim is to empower the user and offer him a choice of ISPs. In addition to choice, we also want to provide widespread access to the Internet. No longer does a person have to be confined to the four walls of his home, office or school to have Internet access. He should be able to sit in a park, a café, a train station, and even on an island and still have access to the Internet. As mentioned earlier, we have chosen IEEE 802.11b wireless LAN technology for this business model. However, our model makes room for the adoption of and migration to newer, better technologies as they are developed and commercially introduced (HiperLAN/2 and IEEE 802.11a for instance).

12.1 SwedenOpen network – A brief description
SwedenOpen (SO), an independent entity (composition described later), will set up and maintain a network consisting of access points all over Sweden, mainly in densely populated areas. These access points intend to cover public places such as Museums, Parks, Libraries, Parking lots; commercial establishments such as hotels, cafeterias, and restaurants; and also schools and colleges. All access points will be connected to the SwedenOpen Internet eXchange (SO IX). The SO IX in turn will be connected to the networks of various participating ISPs, who will bring their cable to the IX. A customer will go to a location that he knows is covered by the SO network. There will be a map on our32 website describing where the Access Points are. He will be able use his Internet enabled device and access our local site that is free for everybody to use. This page is customized according to the location of the user. If the user is at the City Hall, this page will give him information about the City Hall and surrounding places. The user can choose any ISP if he wants to browse the Internet.

32

“Our” means of SwedenOpen. In this model, we consider ourselves to be the entity named SwedenOpen.

54

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Telia Bredband

End user Sweden Open IX Bredbands bolaget

Internet

SwedenOpen Access Point

Content WebServer

Song

Figure 6: Layout of SwedenOpen Network

12.2 End-user experience
There are many ways in which a user can use the SO network. The reader will get a much better idea about the model as we describe the user experience in this section. A user needs a laptop, PDA or some other Internet enabled device in order to use the SwedenOpen network. Either the user should have an IEEE 802.11b wireless LAN compatible PCMCIA card, or the device should have IEEE 802.11b compatibility inbuilt.33 The user can find a list of all locations that are covered by SO, through the www.swedenopen.net website. Once the user is at a SO location, and opens his web browser, he will automatically be taken to a locally customized version of www.swedenopen.net. For instance, if he is at the City Hall in Stockholm, he will reach www.swedenopen.net/stockholm/cityhall. That page will display local content such as information about the City Hall and surrounding locations – including transportation information, list of restaurants in the area, and other content. This information can be viewed for free. If the user wants upstream Internet connectivity, there could be different scenarios. In order to aid the reader’s understanding of the user experience, we present the steps in an algorithmic fashion:

33

Many OEMs such as Dell, Symbol, etc. are either planning to or have already rolled out devices that have inbuilt 802.11b capability.

55

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

1.

Customer boots up, opens his browser and comes to a Swedenopen.net page; this page is customized according to the location of the user. If the user is at the City hall, this page will give him information about the City hall and surrounding places.

2. 3. 4. 5.

Now, he wants to log on to the Internet. He is asked “Which ISP do you want to use?” He says “Telia”. (He could choose any ISP at this point, but we will use Telia in this example; the sponsor definitely gets the preference). Are you already a member of Telia ISP? a. If yes, i. Log on using your Telia ISP user name and password. (The user will pay extra to his ISP if he wants access to all SO sites; it depends on the ISP how they want to charge the users). b. ii. Browse. If no, i. Would you like to sign up for Telia ISP? 1. If yes, a. Sign up and become member of Telia ISP. Someone who wants to use Telia every time will do this. In this case, he will get a monthly bill from Telia. (The pricing scheme is discussed later.) 2. b. If no, a. Log on and browse. Do you have a SwedenOpen Surf Card? i. If yes, 1. 2. ii. If no, 1. Would you like to buy one now? a. If yes, i. Enter credit card and other information. ii. Get a virtual Surf Card. iii. Enter the access code and PIN number from the card. iv. Browse for the amount of time on the card. b. If no, i. We are sorry. You will have to buy a Surf Card or sign up with an ISP to be able to surf the Internet. Enter the access code and PIN number from the card. Browse for the amount of time left on your prepaid Surf Card.

ii. Give more information about
how the user can buy the Surf

56

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Card from various locations in the city. Also give a phone number that he can call to get more information and potentially to buy a card. (He might be more comfortable giving the credit card information on the phone than on the web.)

Someone who wants to use different ISPs every time is more likely to buy a pre-paid Surf Card rather than sign up with an ISP. This can be bought at all locations, which sell mobile telephone pre-paid cards today. If the user’s Surf Card has reached its time limit, the user will have the option to “refill” his card if he is willing to provide his credit card information over the web. While using a SO Surf Card, any particular ISP does not captivate the user; he has a choice every time he logs in. Each ISP is likely to provide incentives to people to become their captive members by providing attractive rates. Please note that a person could be a captive member of two ISPs at the same time. Being a captive member does not mean that he does not have the choice anymore. It only means that he will have to perhaps pay a monthly bill to his chosen ISP i.e. Telia in this case. Once the user has chosen an ISP and is browsing the Internet, all the advertising revenues about local businesses, etc. would go to the ISP, and SO has no ownership of that revenue.

12.3 Roles and Value Chains
In this section we will discuss the value chain and the different roles of each player in the chain. Figure 7 displays the value chain of SwedenOpen. The arrow direction indicates the flow of money through the network. Please note: All the financial figures mentioned in the remainder of this section are backed up a detailed financial analysis presented later.

12.3.1 SwedenOpen
SwedenOpen is an independent private corporation, which acts as the hub of this system. It is a profit making entity. The composition is discussed in more detail later. a. SwedenOpen (SO) is responsible for laying the cable from the Internet eXchange to the Access Point. It will buy or lease the cable from an infrastructure provider such as Stokab in Stockholm. b. SO is also responsible for setting up all the Access Points and maintaining them. It will actually outsource installation and maintenance to a third party. We expect the total cost of hardware and installation to be about 3000kr per access point, not including the cost of leasing the cable.

57

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

c. SO is responsible for customer support. The customer will get a toll free phone number to SO. If the problem is SO related, the customer service person will solve it. If it is related to a particular ISP, SO will forward the call to the customer service representative at the concerned ISP. For the customer, the transfer will be seamless; he will not have to dial again to get to the ISP customer service. Again, the SO customer service be outsourced to a third party.
Advertising Revenues

ISPs (Telia etc.)

Hardware Vendors

End user HardPublic Locations ware

Sweden Open

Local Web Content Provider Installation, Maintenance and Customer Service

Surf card Cable Provider Vendors

Figure 7: Value Chain for SwedenOpen

12.3.1.1 Value Proposition for SwedenOpen
• • • Being the hub of this model, SwedenOpen will control the flow of money and services. SO will be able to generate revenue through several sources including revenue sharing with ISPs and locations, and advertising revenues. SO will pioneer the first operator neutral wireless network in the country, and probably, in the world. Based on their experience in Sweden, they can extend the model to other parts of the world. 58

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

12.3.2 Infrastructure Provider
An infrastructure provider will build and run a fiber-optic network that maintains consistently high security and accessibility. It will function as an operator-neutral provider. The infrastructure provider will lease the fiber to SO for establishing connections from the SO IX to the Access Points. An example of such an infrastructure provider would be Stokab in Stockholm.

12.3.2.1 Value Proposition for Infrastructure Provider
• • • Sell more fiber and make money. The infrastructure provider will get a big order because SO will cover several locations within a particular geographical area. Possible equity stake in SwedenOpen as discussed later in Section 12.4.

12.3.3 ISPs
a. Each ISP will have to bear the cost of getting their cable up to the IX. b. SO will offer the ISP two different kinds of deals and the ISP can choose whichever it wants34: i. The ISP pays a monthly rent for using the IX, and collects all the user revenue generated through its network. It does not have to share revenue with SO. ii. It does not need to pay a monthly rent but will share revenue with SO. This will approximate to a 50-50 split. c. All the advertising revenue generated after the user logs on to the network of the ISP will go to that ISP and not to SO. d. The ISP will have to sign zonal agreements i.e. if the ISP signs an agreement for the Stockholm area, then it has to participate in all locations within Stockholm – this is done to allow a uniform customer experience within a given geographical location. An end user would not want to see that Telia ISP is available as a choice when he is surfing at the City Hall, but is not available at the Vasa Museum. However, Telia would have the option of staying out of the Gothenburg area completely. Thus, either an ISP has be present in the whole zone, or stay out completely. e. The ISPs will have full freedom to decide the price they want to charge the end users. This will allow them to give attractive offers to the customers, and entice them to join. The interaction between the ISP and end user could take place in several ways: ! The end user is already a customer of the ISP’s wireless or wire line broadband offering; the customer has to pay extra in order to be able to use the ISP’s network through SO. This extra charge will be 300kr per month.

34

Disclaimer: We are recommending two feasible types of deals here, but other options will also have to be kept open by both sides when the actual discussions take place.

59

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

!

!

The end user signs up as the member of a particular ISP just for access through SO. The user could do this by calling the ISP or “on the spot” at a SO location through the Internet. He will receive a monthly bill from the ISP. In this case, the ISP will charge 600kr per month. The end user buys the Surf Card and uses the ISP’s network through SO. In this case, there is no direct relation between the ISP and the customer.

12.3.3.1 Value Proposition for ISPs
! SwedenOpen will allow each ISP to extend their service to areas where they would not have been able to reach on their own. Using SO, the ISP just needs to bring its cable to the IX and it will automatically get access to consumers all over Sweden. It will be much easier for the ISP to just bring their cable to the ISP than having to cover the entire city in one way or another. Some ISPs might see this model as a path to commoditization of their service. But we believe that this will be a blessing in disguise for the ISPs as increased competition will foster innovation and development of value added services.

! !

12.3.4 Hardware Vendors
These are players like Ericsson who would supply Access Points and Wireless LAN cards. SO will invite proposals from various hardware vendors for a bulk order of access points and wireless cards. (For example, 5,000-10,000 access points and cards over a period of 3 years). The SO will pay hardware vendors for their equipment. Since this will be a bulk buying deal, each access point would cost anywhere between 1500-2000kr.

12.3.4.1 Value Proposition for Hardware Vendors
! ! ! Sell access points and wireless cards and generate revenues; a long-term deal will give guaranteed revenue over several financial quarters. This would be a very prestigious deal for the hardware vendor leading to sales from other avenues. The hardware vendor might be interested in having an equity stake in SO in return for hardware.

12.3.5 Locations
As mentioned earlier, SO will install Access Points at various locations including public places, commercial establishments, schools and universities. Let us explore some of them in detail: a. Commercial Establishments - Privately owned places like hotels, restaurants, cafeterias, etc. will be very attractive locations for SwedenOpen because people spend a lot of their

60

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

time in these locations and would like to have access to wireless Internet in these places.35 SO will offer two deals to these establishments and they can choose any one of them – ! The location pays for the hardware equipment and the monthly broadband charges; and the revenue that is generated is shared among the location, SO and the ISP. The locations will get 10% of the revenue generated by SwedenOpen. The location does not pay for hardware or broadband access, but they do not get a part of the revenue generated. SO might pay them a small sum of money upfront in order to be able to use their location.

!

We think that the first deal will be a better one for everyone because the location providers are much more likely to promote the service if they have a vested interest i.e. revenue sharing. However, some locations that are skeptical of the whole idea would like to opt for the second deal, and once they are convinced of the value proposition, they will be allowed to shift to the first deal. In any case, the locations will be allowed to show their content on the initial free pages. The advertising revenue from these pages will be split between SO and the location. The value propositions for the commercial establishments would be an increase in the number of customers36, revenue sharing with SO and improved service to their customers, thus generating a positive word of mouth. b. Archipelago Foundation – About 20 islands in the Swedish Archipelago have access to fiber that Stokab has provided. The main reason for the cable to be there is that a lot of companies, like taxi companies and various customer support businesses, have relocated their communication centers out in the Archipelago to avoid the absurd rental costs in the main city as well as to provide work for people in the middle of nowhere. However, this is not generating enough traffic on the broadband network and hence Stokab would like to make better use of the fiber. Wireless access will be provided at hotels, conference centers, and “boating hotspots”popular harbors where people get together for boating and where a lot of people live all round the year. SO will strike similar deals with the hotels and conference centers as described in Section (a) earlier. For common public places like harbors, the Archipelago Foundation will bear the cost of hardware installation – they would do so in order to attract

35

Based on the results from the End-user survey that we conducted in the US and Sweden, more than 25% of the respondents spend more than 10 hours at a café every month. It is also validated by the fact that most of the early players in this market are targeting these locations. Most of the service providers we interviewed expressed the same opinion as reflected by our survey results. 36 According to the End-User survey, more than 45% of the people responded that they would choose a café over another one based on the fact that the first one provides wireless Internet access.

61

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

more people to come to the islands, and as a service for the tourists. If the Foundation does not agree to pay for the hardware, SO will bear the cost itself. c. Public Places – Public places such as parks, museums, train stations, subway stations, libraries, etc. are controlled by the government. These locations also have a tremendous potential for generating wireless Internet usage. SO will first try to convince these authorities to bear the cost of installing hardware as a service to their visitors. If such a deal does not work out, SO will have to offer them the two deals that we mentioned in Section (a). d. Schools and Universities – The Internet revolution was born in universities and other academic institutions. In what has been seen till now, universities have been the early adopters of wireless LANs – IEEE 802.11b in particular.37 SO will establish relationships will all the major schools and universities in Sweden. The universities will pay a monthly fee to SO to manage the wireless Internet access. The students will have the option to choose their school as one of the ISPs on the SO network. They will be able to surf for free as long as they are using their school as the ISP. This free access will only be available while they are at school, and not outside school. People from outside who are visiting the school will be able to get onto the SO network and choose their preferred ISP – but service will not be free for them. We think that the students will not be interested in choosing any other ISP because they get free access through their school. However, in the future, as ISPs focus more on value added services, students might be willing to pay for gaming, movies or something similar from a specific service provider. The value proposition for schools and universities is to provide wireless Internet access to their students, faculty and staff and in the process fostering new research in this area. Also, they would like to provide access for outside people visiting the school. e. Homes – SB (Svenska Bostäder) will run the cable from the SO IX to all their newly built buildings and also to all the buildings that they remodel. However, buildings and homes already connected to a particular ISP and have a multi-year contract will not be able to participate in the SO network. If the building is connected through SO, the user will get the choice in wired Internet access. However, if she wants wireless access in her home, she will have to put up an Access Point on her own, and unlike the Share Point model, it will not be managed by SO. SO could take up that role in the future – but it should try to focus right now on wiring all the other locations mentioned above and driving usage at those locations.

37

KTH in Sweden; and Carnegie Mellon University in the US are examples.

62

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

The value proposition for SB is that they will be able to offer a choice of ISPs to their prospective customers, and thus make their housing schemes more attractive. They will also be able to charge a premium for providing this facility. SwedenOpen should try to sign exclusive contracts with all locations in order to gain a sustainable advantage, and lock out competition. However, if that is not possible, competitors might have their base-stations in the same location too. If a competitor wants to set up a base station at a location already covered by SwedenOpen, SwedenOpen should “sublet” their network to the competitor so that the competitor does not have to set up access points and customers of both providers can access the same network. The other provider should pay a sublet fee to SwedenOpen. In case, SwedenOpen wants to cover a location already covered by a competitor, they should also follow the same procedure and pay a fee to the original player. Such a method will prevent redundant networks and interference.

Park Parking Lot Condos Airport Bus station

Train station Archipelago

Library / Museum

Stadium
School

Figure 8: Locations covered by SwedenOpen

12.3.6 End users
We have already talked about the end user experience in Section 12.2; let us now talk about the different profiles of users: a. Already use some ISP (wired) – If the user already has access to Telia ISP at home, then he can use Telia through SO and the additional charges will just show up on the monthly 63

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Telia bill. Alternatively, Telia can float an offer such that the customer can use any of the SO sites (using Telia) for an unlimited time by paying an extra 300kr every month. If this person has Telia ISP at home, and wants to use some other ISP through SO, he will have to buy a Surf Card. This will be done to avoid sending him 2 bills and to minimize the work for all the parties involved. This process could be further simplified in the future if the ISPs sign roaming agreements with each other – we do not expect that to happen during initial phase of SwedenOpen because of lack of sufficient usage. However, at a later date, the ISPs could sign roaming agreements with each other; in that case, if a person has Telia ISP at home, and he wants to use some other ISP through SO, the additional charges will show up on his monthly Telia bill and the ISPs will settle accounts amongst themselves. b. Use a competitor like Telia Homerun – A Telia HomeRun user will be able to use the SO network if there is a roaming deal between SO and Telia HomeRun. In this case, the user will be asked an additional question while using the service “Are you a current member of the Telia HomeRun service?” If the user answers positively, he will be able to enter his Telia HomeRun username and password, and choose any ISP to access the Internet. If they have a roaming deal, the charges for using the SO network will show up on the user’s monthly bill from Telia HomeRun. c. Both (a) and (b) – If there is no roaming deal between HomeRun and SO, then this case is similar to case (a). If there is a roaming deal with HomeRun, the user will have a choice of either getting charged on his HomeRun bill or his Telia bill. If he plans to use Telia as the ISP every time, he is more likely to pay a fixed additional monthly charge to Telia. However, if he wants to use a different ISP each time, he will more likely charge it on his HomeRun bill, thus not having to buy the Surf Cards. We should note that he might have to pay more if he chooses to pay through Telia HomeRun depending on the nature of the roaming deal. Since Telia HomeRun is such an expensive service today38, they will be reluctant to sign a roaming deal with SO on cheaper terms fearing a diminishing use of their own network. d. Use nothing right now – In this case, the user will either become a member of a particular ISP, or use the Surf Card system. The user could buy a surf card for $3 to user for an hour, $5 for 2 hours, and $10 for 5 hours.

38

They charge about 1500kr per month.

64

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Value Proposition for the End User • • • • As discussed earlier, the end user will get a choice of ISPs and will not be forced to use a particular ISP. If she already is a subscriber of a particular ISP, she could get her SO charges added to the same bill, thus reducing paper work. She has the option of trying out the service for less by buying the Surf Cards, which would be available at various outlets. She will have Internet access at places like the Archipelago, parks and cafeterias – something she could only dream about till now.

12.3.7 Local Web Content Provider
SO will outsource the development of the local web pages to a small website design firm. Alternatively, they can also hire 2-3 web designers to develop the web pages. Most of the locations will already have their home pages. For instance, the Vasa museum can be found at http://www.vasamuseet.se. Locations that do not have homepages will be requested to develop their own pages, which would help them popularize their location. SO or the design firm will then customize these homepages in such a way that the SwedenOpen brand is visible on all these pages in a consistent manner. If the location does not wish to develop its homepage, SO will display a generic page at that location, with information about surrounding places. The selling and display of advertisements will be controlled by SO and all the advertisement revenues from these pages will be owned by SO.

12.3.8 Surf Card Vendors
The refillable surf card will be for sale in locations where SO has an AP. It will also be for sale everywhere where you today can buy refillable card for mobile telephones. The idea is to make these cards very accessible for people – they would be available at train stations, 24-hour convenience stores, supermarkets, etc. In fact, SO will avoid having to establish a distribution channel for the cards. They could just have an agreement with the distribution companies for mobile phone cards, because they have an established footprint in the same target market. The vendors will make a small margin (approximately 5%) on the sale of these cards – the margin will not be very high because these cards do not occupy much shelf space. The distributor will make a margin of about 15%.

12.3.9 Advertising Revenues
All the companies that choose to advertise on the local pages will be charged based on the number of impressions of their advertisements. They will be charged 100kr per thousand impressions – this rate might be a little higher than market rates, but these advertisements will be targeted locally and hence highly effective. All the revenue from these advertisements will go to SO.

65

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

12.4 Roadmap
It is clear that in order to make this model successful, SwedenOpen needs to work on many fronts and establish relationships with several players. The resources (labor, time and money) are limited and will ramp up as time goes by. Also, SwedenOpen needs to give a proof of concept in order to get a universal “buy-in” from all players in the value chain. The good news is that KTH and Stokab have already started a small-scale project in Kista and other parts of Stockholm under the auspices of StockholmOpen.net. The success of this project would provide a great proof of concept for SwedenOpen. We feel that the most important first step is to establish the composition and ownership of SwedenOpen. As stated earlier, we assume that SwedenOpen will be a for-profit venture run by entrepreneurs with the ultimate aim of having an Initial Public offering. Potentially, Ericsson Business Innovation could “spin out” SwedenOpen providing it with an experienced operational team. Ericsson Business Innovation would take an equity stake in the company in return for capital.39 Ericsson could also leverage this investment to sell hardware to SwedenOpen. In addition, an advisory board will be created consisting of people from KTH, Stokab, and Svenska Bostäder who are already working on StockholmOpen.net and can provide valuable guidance and resources to SwedenOpen. These people will also have an equity stake in SwedenOpen irrespective of the fact whether they invest capital or not. If SwedenOpen can convince these people to invest capital, that will be a great early victory for the concept, and a guaranteed source of advice in the future.40

39

Ericsson Business Innovation has a venture model based on the principles of early business focus, encouraging entrepreneurial spirit and capitalizing on created values – from ‘www.ericsson.com/innovation’ 40 Please note that only individuals would form the advisory board and whole organization such as KTH or Stokab will not have a direct investment or an operating role in SwedenOpen. SO will be a totally private corporation unless it declares a Public Offering.

66

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

After the management structure has been put in place, SwedenOpen will start executing on the following milestones: Confidential

67

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

12.5 Sustainable Advantages of SwedenOpen
A successful business model requires sufficient barriers to entry for competition to come in. If the competition is allowed to enter the market too soon, the value of SwedenOpen will erode in no time. We believe that the model as described above will create several sustainable advantages for SwedenOpen: ! An experienced board of directors (including Ericsson), and a very knowledgeable Board of Advisors who have proved the concept through StockholmOpen.net. With such big names behind them, SwedenOpen will be able to generate unparalleled interest from ISPs, locations, and end users. We foresee the wireless LAN market as a land grab opportunity for the next 2 years. By partnering with a wide range of locations quickly, SwedenOpen will leave little space for the competition, and will be able to draft any roaming agreement according to its terms. Once a location becomes a “SO location”, the switching costs are very high, especially if they sign a 2 to 3 year contract. SwedenOpen will own patent rights over any technology that is developed as a part of the SO IX; this will make it very difficult for competitors to replicate the model.

!

! !

12.6 Risks and Challenges 12.6.1 Technical Implications
All the technical implications listed for the “Share Point” model under Section 11.6.1 will also be applicable to this model because both are initially based on IEEE 802.11b technology. In addition, the following technical challenges exist in this model: In order to divide the traffic according to the users choice of ISP, a traffic-handling server is needed in SwedenOpen. If a user has chosen Bredbandsbolaget as his ISP, but requests a page from a Telia server this traffic handling server makes sure that the user is still surfing and gets a bill from Bredbandsbolaget.

68

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

4 Bredbandsbolaget 1. The user chooses Telia as his ISP. 2. The traffic-handling server notices that this user wants to use Telia. 3. All Internet services are handled via the IX with Telia. 4. The user requests a page from a server at Bredbandsbolaget. 5. The traffic handling server makes sure that the traffic is still routed via Telia and that the end user doesn’t get billed by Bredbandsbolaget. 1 Song Telia ISP

3

Internet eXchange Sweden 2 Open

Traffic Handling

AP

Figure 9: An illustration of the traffic-handling server. !

!

IXs are the bottleneck of the Internet traffic. However, in order to combat that, the Swedish ICT Commission has announced a vision of one IX per every 30,000 households in Sweden. This could lead to subsidizing from the government, which is positive for SwedenOpen. A very intelligent and extensive billing and authentication system will be required to distinguish between people using various payment methods to surf.

69

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

12.6.2 Market Risks
Potential Risk41 ISPs consider SO to be a threat that might lead to their commoditization; only a few ISPs sign up. Pre-emptive Move Convince them that SO will allow them to reach locations that they would have never been able to reach otherwise; sell them all the value propositions discussed in Section 12.3 Balance the coverage between indoor and outdoor locations Corrective Action Run the pilot program with a few ISPs; the rest will join after seeing the success of the Pilot.

Usage is almost nil in many public places during the long winter months in Sweden Funding is not easily available because of tight economic conditions

A fur coat free with every Surf Card ☺

Build a great management team, and Run the pilot program by asking the City to fund it. Board of Advisors that will lend It will be easier to raise credibility to the company capital after demonstrating a successful pilot. Sell wireless cards to the customers of SO at a small subsidy by buying them in bulk from Ericsson Target ordinary users and students, and not just business users Start renting out wireless cards at various locations that are covered by SO42 Target other countries with a similar model; especially developing countries which do not have good wire line infrastructure. SO should pay for setting up the hardware and also pay a small upfront fee to the location owners. HomeRun is targeting business users and charging a huge amount; undercut their price to acquire users.

Users do not have wireless cards, and are unwilling to buy them Swedish market is not big enough to create economies of scale

Location owners are not tech-savvy enough and refuse to set up access points in their premises Telia HomeRun and other competitors gain sufficient ground because of their first mover advantage; capture large market share
41

Show them examples of other places that have installed the access points and thus increased customer flow. Tempt them with the revenue sharing deal. Strike a roaming agreement with HomeRun to create a synergistic partnership

This matrix has been adapted from the DDART framework taught by Prof. Tom Kosnik, Consulting Professor, Stanford University in his class on Global Entrepreneurial Marketing. 42 This model is covered under the section “Other Ideas and Recommendations”.

70

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

A member location wants to also allow a competitor of SwedenOpen to put up their base station at their location.

Try to sign an exclusivity agreement Place access points strategically to avoid with the location. If they don’t interference. agree, recommend them that it will be better if they do not have other providers in the same place because it might lead to interference between access points. Table 5: Market Risks for SwedenOpen

12.7 Financial analysis
Please refer Section 20.1 Confidential for the assumptions used, and Section 20.2 Confidential for a detailed income statement for SwedenOpen over the next three years. According to our calculation, SwedenOpen will become profitable in Year 3 making a profit of 44 million kr with total revenue of 262 million kr.

71

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

13 OTHER IDEAS AND RECOMMENDATIONS
Our research and brainstorming yielded several ideas, a few of which developed into business models presented in the previous section of this report. However, we believe that some of the ideas that did not end up being a part of the business model for several reasons43 still bear a lot of merit and could prove to be very useful when developed further. Therefore, we feel it is highly important that we share them with the readers. Please note that the ideas may vary from being moderately thought-out models to completely wild flashes of thought – both of which are essential parts of creative brainstorming. We can divide these ideas and recommendations into three different sections:

13.1 ‘Add-on’ Ideas
This section talks about ideas that can be incrementally added to the business models described in the previous section:

13.1.1 “Rent-a-Sony” Model
When a user enters a location that has wireless Internet service available, he may fall into one of these four categories: No Service Has Device Has service Has device

2 3 1 4
No Service No device Has service No device

By “Device”, we mean a laptop or a PDA that he can use to access the Internet. Currently, the wireless access providers are trying to convert people from Category Two to Category Three. There are 50 million people in the world that have laptops and the number is expected to increase to 100 million by 2003.44 Therefore, the access providers are focusing on this segment of the population as their target market.

43

Reasons include constraint of time, dependence on unproven technologies. We want our current business models to be deploy-able immediately. 44 Nokia Operator Wireless LAN Solution Brochure

72

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Category One is the biggest category of people consisting of the rest of the world population. There is about 428 million mobile phone subscribers today, expected to grow to 1 billion by 2004.45 We believe that at least half of these subscribers are also potential customers for wireless LAN. They want to access the Internet in public hotspots but do not want to buy a laptop, which costs upward of $1000. If we target this market, the potential market size for wireless LAN users suddenly jumps from 100 million to 500 million in 2004. This can be a very lucrative market if addressed appropriately. We recommend that hardware devices should be made accessible to people at public hotspot locations. When a user enters an airport, and finds out that he has a 2-hour long wait before the plane takes off, he should be able to rent a laptop or a PDA at the airport, buy service on the spot and surf the Internet.46 The same applies for cafeterias, shopping malls, hotels and other places – but airports have the maximum value proposition. Compaq has recently struck a deal with Starbucks in which each Starbucks store will have Compaq iPAQ handhelds to be rented out to their customers for the period while they are at the store.47 There are several new partnership opportunities for Ericsson or Telia in this model. Ericsson can leverage their recent joint venture with Sony to make Sony Viao laptop computers and Sony Clie handhelds available for rent at airports and other places. We believe that Sony will be extremely enthusiastic about the visibility their products will get in the US and European markets. Also, Palm and Handspring will be ideal partners since they would love to replicate what their competitor Compaq has done. In order to make this model possible, every location needs a distribution outlet. We feel that building a new outlet is not a scalable solution. Instead, we should try to partner with already existing franchises at these locations. For instance, Internet kiosks are an emerging market at several airports and other locations where people are forced to wait such as hospitals, malls, hotels, etc. Ekiosk, headquartered in New Lenox, Illinois, already has about 1000 Internet kiosks in the US, and plan to have about 10,000 by the end of 2001.48 Ekiosk would be an excellent distribution partner due to their extensive national footprint. The wired Internet kiosks might consider wireless LAN providers as their competition. But we feel they will realize that wireless LANs are a very strong phenomena, and it would be wise for them to embrace the technology and create a whole solution at their Internet kiosks by providing both wired Internet and wireless Internet. These outlets should also carry wireless cards to serve the consumers with or without laptops.

45 46

Nokia Operator Wireless LAN Solution Brochure “In the year 2000, one out of every four flights was delayed in the US, affecting 163 million passengers.” www.cnn.com May 10, 2001. 47 http://www.compaq.com - Press Release on May 2, 2001. 48 The Standard, “Peace, Quiet and a T-1 Line”, May 3, 2001.

73

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

Though this model entails increasing the number of stakeholders in the value chain, a premium can be charged for renting out the devices, thus creating value for all the involved parties. We recommend that Ericsson and Telia develop this model further and explore partnerships in this area.

13.1.2 Roaming agreements
The strength of the network increases exponentially with every additional node. This famous adage holds true for a wireless LAN network – more the number of hotspots, better is the value proposition for the user and greater the revenue opportunity for the provider. However, installing wireless LAN access points is a capital intensive and time-consuming activity, and we believe that no single operator can create a network in every part of a country, and the world. Therefore, we recommend that Share Point or SwedenOpen enter into roaming agreements with local, national and international access providers. This will allow users of these services to travel nationally and internationally and still have access to the Internet using their provider back home. Since Telia have experience in forming roaming partnerships in their mobile phone business, they can leverage that knowledge to create similar partnerships in wireless LAN networks. HereUare (described in Section 9.5) is a company that enables roaming between different wireless service providers and might be a good partner for outsourcing roaming solutions. Some potential roaming partners that we feel can add value include: • • MobileStar – United States Wayport – United States

• SkyNet Global – Australia Efforts should be made to explore other roaming partners in other parts of Europe.

13.1.3 Seamless Handover
One of the major limitations of wireless LANs (by their inherent nature) is the limited range – it does not allow a person to move around freely beyond a certain limit. We agree that people will not want to browse the Internet on their laptops while walking around. However, PDAs have the appropriate form factor for being used for that purpose i.e. browsing while you are standing on the road, walking on the beach, etc. In that regard we feel that LAN technologies could be complimentary to upcoming high-speed WAN technologies such as GPRS and 3G. Currently, there are no solutions on the market that allow a person to seamlessly move between a LAN and a WAN environment without losing the connection and having to reconfigure. However, in the future, Ericsson and Telia should explore solutions that can solve this problem. Some companies such as ColumbiTech and Netseal are trying to develop a solution that enables seamless roaming between different networks. Such a solution would be a great value proposition for the users, and Telia could combine their WAN and LAN solutions to charge a premium, thus increasing their Average Revenue per User. 74

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

As an example, a Share Point customer who is already a Telia ISP customer will be more interested in becoming a Telia Mobile customer because he would be able to seamlessly roam between the Share Point and Telia Mobile networks.

13.1.4 Upgrade path to IEEE 802.11a
As mentioned earlier, IEEE 802.11b uses the unlicensed 2.4 GHz frequency band. Because it is unlicensed, anyone can set up an access point and broadcast on the frequency. However, there is a limitation to the usage of the frequency in one location, interference will occur if the frequency band is cluttered. No more than three access points can be put in the same place, as there are only three available channels for transmission. Also, the 2.4 GHz band is cluttered with several other interfering devices such as the microwave, cordless phones, and Bluetooth devices in the future. On the other hand, the next generation IEEE 802.11a technology uses the 5 GHz frequency band, which is much less cluttered and provides 12 channels for transmission. Therefore, the Quality of Service will be much better than with IEEE 802.11b. We recommend that Ericsson should partner with companies such as Atheros Communications and Radiata that are trying to develop chipsets for IEEE 802.11a-based devices. Though Ericsson was late to the party in coming out with IEEE 802.11b based hardware, they can redeem their market share and lost profits by taking the lead on IEEE 802.11a compliant hardware. Share Point should be aware of these future developments and create schemes where users can trade-in their old IEEE 802.11b wireless cards for IEEE 802.11a cards. Also, the access points should be easily upgradeable.

13.1.5 “Grocery Shopping” model
Imagine a Share Point or a SwedenOpen customer shopping groceries at a huge supermarket. If the supermarket is covered with Access Points, the user can pull up her49 wireless enabled PDA and browse the homepage of the supermarket. She can find the aisle number for the product she is looking for, and also receive targeted promotions from the supermarket. She could have also saved her shopping list on the homepage of the supermarket and can access it for free from her PDA now. If she wants to get connected to the Internet, she can just use her username and password and surf. We feel that this is a highly value added service for both users and retail stores – increasing sales and ease of shopping. Taking it further, the supermarket can use the same network for taking stock of their inventory thus enabling error free and quick stock taking, further leading to real time ordering of products, and also connection with their backend systems. This is an example of being able to use the same network for customers and employees. However, there are security concerns with the current IEEE 802.11b standard which might prevent stores from using this network for their mission
49

No gender bias intended. Men can go grocery shopping too.

75

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

critical and sensitive activities. But we expect that the security concerns will be mitigated in the future and such applications will become more practical and widely accepted.

13.1.6 Bundling hardware with service
In order to sell more hardware, Ericsson should partner with Telia and bundle their hardware with the service offerings that Share Point or SwedenOpen will provide to their customers. The hardware can be subsidized for the users who sign up with either of the services – a model similar to the one followed by mobile operators today. This will reduce the cost for the end user who would otherwise have to buy the hardware and the service separately, and increase the revenue and profits for Ericsson, as they will save on distributor and retail margins on their products.

13.2 Potential Business Models
This sections briefly mentions ideas that could be seeds of new business models50:

13.2.1 Vertical Markets
Each vertical market such as automotive, retail, healthcare, hi-tech or hospitality has its own backend issues including inventory management, warehouse management, logistics and supply chain management. Companies are using wired local area networks or proprietary wireless networks to co-ordinate such activities. We feel that IEEE 802.11b based networks can add many values such as ease of installation, ease of use and a clear upgrade path to future technologies. To make things more clear, let us take an example. Imagine a consumer good manufacturer such as Sony, having a warehouse for Sony Televisions in San Diego. They can use IEEE 802.11b compliant wireless bar code scanners to take stock of the inventory and it is transferred in real time to their backend system. Symbol is already developing such systems, and could be a potential partner. Telia and Symbol have already partnered in a “Wireless LAN enabled Ambulance” project in Sweden.51 There are many other interesting applications of this technology in backend logistics.

13.2.2 In-transit wireless Internet
More than 50% of the respondents to our end user survey both in the US and Sweden said that they would find Internet access “very useful” while traveling in cars, trains and airplanes. In fact, airplanes and trains were by far the most preferred locations for access in the US and Sweden respectively. During our “on the road” survey in the US, several people expressed a strong desire for Internet in cars. We feel this is a tremendous opportunity, which is not being addressed right now. Some airlines are carrying out trials to provide in-flight Internet service, but mass deployment is still far off. We recommend that Telia and Ericsson undertake a joint initiative to

50

We are not going into details of these models, however we will be happy to brainstorm about details during the presentation or any other time. 51 http://www.nwfusion.com/reviews/2001/0205bgside.html “Saving Lives with Roving LANs”, February 5, 2001.

76

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

address this obvious and huge market. Tenzing Communications, a Seattle, United States based company is developing solutions to provide in-flight Internet access, and might be a potential partner. The train authorities in Sweden and the major auto manufacturers of US and Sweden could also be potential partners as they have a great value proposition in making this possible.

13.2.3 Going International
Ericsson and Telia should explore other countries as potential markets for wireless LAN networks. Different countries have different regulatory and socio-economic environments, thus requiring different business models. We feel that they should first focus on other European countries, and then the Asian markets – starting with more developed countries like Singapore, Hong Kong, Japan, etc. A company named Wemobile is planning to deploy an IEEE 802.11b based network in Singapore as the basis for a pervasive (not just hot-spot) mobile broadband service.52 They expect that blanketing Singapore would require something in the order of 50,000 base stations. This might be a great sales opportunity for Ericsson and later on, a potential roaming opportunity for Share Point and SwedenOpen.

13.3 General Recommendation
According to our survey, 96% of the people would like to have Internet access away from their desk or home, and at least 90% are ready to pay for it. There is no mistaking the fact that WIRELESS INTERNET ACCESS IS A HUGE OPPORTUNITY. We recommend that Ericsson and Telia give strategic importance to this market. The competition is getting tougher each day; they need to move fast in order to establish their leadership in this market, in their respective roles. We wish them the best of luck!!

52

http://www.wemobile.com

77

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

14 FURTHER RESEARCH
We recommend various areas for further research if any of our business models are to be pursued seriously. In our work, we have understood the capabilities and limitations of the technologies but have not constructed any technical specifications for how the actual solutions will look. Such technical work along with field-testing of proposed equipment and services would be the next logical step before a commercial business can be launched. Furthermore, our market data has been gathered from widely available sources and the costs have been calculated from retail prices. Fixed and other cost estimates involved in the operation of such a venture are based on best knowledge rather than from the experience of actually running a service provider company. Hence we might have overlooked cost centers or overestimated others. An appropriate financial analyst team will have to survey the market, get price quotes and build a more robust and reliable model charting a possible path to profitability. Also, the possibility of acquiring one of the cash strapped wireless LAN service providers in the United States should not be ruled out as this could reduce development costs for backend software and increase time to market with an increased base of deployment experience in the space. To this end, our initial description of the company’s should be used for further due diligence to explore that opportunity. Finally, with regard to some of the ideas and suggestions given above, it is important to keep in mind that IEEE 802.11b will eventually have to be replaced by other technologies or get integrated with them. There must be heightened research into what would be the ideal platforms for this and the development of hardware and solutions specific to these purposes. In addition, we think that all the points mentioned under “Other ideas and recommendations” should be pursued further.

78

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

15 SOME PARTING THOUGHTS
15.1 Were the goals of the project realistic?
Overall, the goals of the project were realistic. The constant interaction with the sponsors and their continuous support in terms of time and project resources helped us in moving towards our goal. We discussed the state of the project in detail with the sponsors at each meeting and reached an agreement on what should be done further. Hence, although at times, the goal was a moving target, we were able to quickly refocus and move towards it. A case in point is that during the development stage of business models, the sponsors reviewed our business models, suggested that we develop two business models – one operator dependent and one operator independent clarified their definition of these terms and thus enabled us to deliver business models that matched what the sponsors expected. However, some of the sub-goals set by the team were unrealistic. One was the time allocated to developing the business models. Another was the time dedicated to writing the report. In both cases, it was strongly felt that more time should have been budgeted for these tasks.

15.2 If we were to re - do this Project, we will …
… meet at the beginning of the project. The mid-term meeting gave a huge impetus to team building efforts and brought the team members very close together. After the meeting, there was a greater sense of camaraderie and a stronger sense of purpose within the team. The team felt that the pace of the project would have been greater if we had met each other at the beginning of the project. … start on the project report and business models earlier and document things in a format suitable for incorporating into the final report. As mentioned earlier, we felt that the time allotted for the business models and the project report write-up was not sufficient. It was also felt that the documentation of the research undertaken throughout the project should have been done in a format that would have allowed us to put it in the final report without making too many changes. … have an undertaker (team leader on a rotational basis ) from the beginning of the project. In our team, we adopted a completely flat organizational structure throughout the first half of the project. Two people (one from Stanford and another from Sweden) were in charge of each task required to complete the project; they subdivided these tasks and delegated them to others and supervised them. However, this turned out to be ineffective. After the midterm meeting, we adopted a form of rotational leadership. It was implemented thus: At the end of every weekly meeting, a to-do list is drawn up based on the deadlines coming up in 79

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

the following weeks and the discussions in the teleconference. One person is designated as the ‘undertaker’ each week at the weekly meeting. Throughout the week, the undertaker checks that the tasks are being completed on schedule. At the beginning of the next meeting, the undertaker reviews the to-do list from the previous meeting and ensures that the tasks were completed. This procedure made the team-members realize that they would be held accountable for their tasks in full public view during the next meeting and motivated them to finish their tasks on time. Needless to say, we feel that this form of rotational leadership should have been implemented from the beginning within the team. … have more meetings with coaches We feel that we should have obtained more guidance from the coaches by setting up meetings with them even if it took considerable effort. By giving them rough drafts of our business models at an early stage, we could have obtained valuable guidance from them.

15.3 Problems faced along the way
! Displacement in time and space Although the time difference can be a blessing when working around the clock, it usually led to considerable problems in contacting team members and coordinating with them on various tasks. Geographic displacement led to difficulty in brainstorming. A solution to this was frequent videoconferences, but the time difference made scheduling these a Herculean task. ! Huge up and down swings in productivity The productivity of the team varied considerably from week to week with bursts of productivity that often coincided with imminent deadlines. It so happened that as soon as we finished something of significance, we entered a state of complacency where things slowed down considerably. This was especially true of our mid-term trip to Sweden. After the trip, the team felt that the project was going well. So, immediately after the mid-term presentation, we lost momentum for a while and it took a very conscious effort to put things back on track. ! Deploying a “perfect” survey! Deploying a survey that would give us accurate results led to us having to delve deep into human psychology☺. Each question was analyzed several times to ensure that it was easy to answer, would not bias the respondents towards choosing a particular answer and would give us information that was relevant to the project. We needed several iterations before we drafted a final version of the survey. ! Socio-economic ignorance. Since both the models are Sweden based, lack of a clear idea about the socio-economic and political aspects in Sweden made it difficult for the Stanford students to undertake a

80

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

comprehensive analysis of these business models. This problem was mitigated to a large extent by frequent conversations between the Stanford side and the Swedish side.

81

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

16 CONCLUSION
The evolution of wireless LANs into consumer markets is definitely the most exciting phenomenon taking place in the wireless Internet market right now. The office desks have seen their Internet revolution, now it’s time to turn the cafés and the airports, the parks and the beaches into Internet “surf-able” zones. The target market for such services is huge, and there is money to be made. Competition is getting stiffer – a land grab is going on. Speed of execution is the call of the hour.

We appeal to Ericsson and Telia that they hear this call, and take the necessary steps to establish their leadership in this fledgling market. We believe that the two suggested models – Share Point and SwedenOpen will be of strategic importance to our sponsors.

Let the games begin!!!

82

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

17 APPENDIX - ACRONYMS AND ABBREVIATIONS
! ! ! AAA, Authentication, Authorization and Accounting. AP, Access Points. CSMA/CA, Carrier Sense Multiple Access with Collision Avoidance, attempts to avoid collisions by using explicit packet acknowledgment (ACK), which means an ACK packet is sent by the receiving station to confirm that the data packet arrived intact. CSMA/CA works as follows. A station wishing to transmit senses the air, and, if no activity is detected, the station waits an additional, randomly selected period of time and then transmits if the medium is still free. If the packet is received intact, the receiving station issues an ACK frame that, once successfully received by the sender, completes the process. If the ACK frame is not detected by the sending station, either because the original data packet was not received intact or the ACK was not received intact, a collision is assumed to have occurred and the data packet is transmitted again after waiting another random amount of time. DSSS, Direct-Sequence Spread-Spectrum technology, avoids excessive power concentration by spreading the signal over a wider frequency band. The transmitter maps each bit of data into a pattern of “chips”. At the destination the chips are mapped back into a bit, recreating the original data. Transmitter and receiver must be synchronized to operate properly. FHSS, Frequency Hopping Spread-Spectrum, spreads the signal by transmitting a short burst on one frequency, "hopping" to another frequency for another short burst and so on. The source and destination of a transmission must be synchronized so they are on the same frequency at the same time. LAN, Local Area Network IEEE, Institute ISM, Industrial, Scientific and Medical, band, this frequency band (2.4GHz to 2.4835GHZ) is a global band primarily set aside for industrial, scientific and medical use, but can be used for operating wireless LAN devices without the need for end-user licenses MAC, Media Access Control MAN, Metropolitan Area Network OFDM, Orthogonal Frequency-Division Multiplexing, is a technology that resolves many of the problems associated with the indoor wireless environment. Indoor environments such as homes and offices are difficult because the radio system has to deal with a phenomenon called "multipath." Multipath is the effect of multiple received radio signals coming from reflections off walls, ceilings, floors, furniture, people and other objects. In addition, the radio has to deal with another frequency phenomenon called "fading," where blockage of the signal occurs due to objects or the position of a device relative to the Internet gateway. OFDM has been designed to deal with these phenomena 83

!

!

! ! !

! ! !

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

! ! ! ! !

and at the same time utilize spectrum more efficiently than spread spectrum to significantly increase performance. PAN, Personal Area Network PTP, Point To Point UNII, Unlicensed National Information Infrastructure, band is open in Europe, the United States and Japan.
WAN, Wide Area Network

! !

WEP, wired equivalent privacy, mechanism covers station-to-station transmission. The standard specifies usage of the RC4 security algorithm. The scheme relies on a 40-bit key to encrypt the payload of data frames. RC4 is a stream cipher designed by Rivest for RSA Data Security. QoS, Quality of Service.

84

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

!

85

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

18 APPENDIX - SERVICE TRIALS
During this course we have performed service trials of wireless Internet access from different service providers. In this section we’ll explain how each service works, problems we faced when we used it and conclusions we’ve made from our experience.

18.1 Sweden
The Swedish part of the team has tried three different networks. They are all based on IEEE 802.11b. Two of them are free to use for students, KTH campus and StockholmOpen.net and one is a commercial wireless network that’s managed by Telia HomeRun.

18.1.1 KTH Kista campus
The entire KTH campus in Kista is covered with wireless LAN access. Every group in this class has been equipped with laptops and WLAN cards. All three students on the Swedish side have been using it since the first day. For both David and Reza, this was the first experience with wireless networks. We faced several problems before we could get everything up and running. First we didn’t have the correct drivers for the WLAN card. Once we got that sorted out we didn’t know what network ID to use (it turned out that any works fine, this way the computer picks any network that’s accessible and uses that.). However, after all the prerequisites got up and running, access has been perfect. We can sit anywhere in school and still be connected to both the Internet and the campus network with printers. Ulrika experienced a different problem. She has been using her laptop and WLAN card both at work and at KTH. The computer doesn’t automatically recognize what environment it’s running in. If you have different settings for work and school you might have to reconfigure your device every time you change networks. This is inconvenient but will hopefully be solved by Windows XP and other new operating systems.

18.1.2 Telia HomeRun
We were equipped with three 24hr cards from Telia. Each one is equipped with a user name and a password that is valid for 24 hrs after the first log on. The price for a card is 96 kronor plus tax. The total with tax comes to 120 kronor for the end user. It requires a device with the capability to use a wireless LAN card for the IEEE 802.11b standard. The installation is very easy. All you have to do is to set the network identification (ESS-ID) to homerun. You need to turn off the encryption settings you might have on your WLAN card as well as proxy settings on your browser. Then you launch your web browser and point it to http://login1.telia.com. Once at the login page you enter the user name and password you just got from the card. Voila, you’re surfing.

86

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

We tried at both Stockholm Central train station and at restaurant Grodan (the Frog) in the middle of Stockholm. Immediately when you point the browser the address above, you are confronted with a login page. After your username has been identified and approved you’re redirected to a location specific homepage called HomeRun@Location, (where Location is changed to wherever you might be). From there you have access to Internet. If you want to access corporate data you can run a VPN-client. A 128-bit client can be bought from Telia HomeRun itself. We thought the service was very easy to use. There are step-by-step instructions on the back of card with username and password as well as a toll-free number you can call for support.

18.1.3 StockholmOpen.net
An operator neutral network called StockholmOpen.net is being launched in Stockholm right now. (Further discussed in the business model section). Currently, in order to access this network you don’t need to be a registered user, all you have to do is find a spot that’s covered. I’ve tried at Kungsträdgården (a big park in down town Stockholm). All I had to do was change my network ID to stockholmopen and point my browser to https://login.stockholmopen.net. At that address I was prompted to enter a username and a password. Since it’s not officially launched, this feature isn’t implemented yet so anything works. On this page I also got to select my upstream Internet provider; only KTH and Bredbandsbolaget were available. Once that’s done, it is smooth surfing.

18.1.4 Problems
If for some reason you want to change your ISP in the middle of a session that is not possible with a logout. You have to remove the card or move out of range from the access point and regain access again. This time you can choose another ISP. There is also no way of knowing what spots are covered. Since this project has just been launched, neither of these services has been implemented yet. However, we talked to the responsible people at KTH about this. They said that it was relevant feedback and that they would implement it.

18.2 USA
The American team was able to try two different wireless networks. Their experiences are described below.

18.2.1 Surf and Sip
We tried the Surf and Sip network at a small café on Polk Street, San Francisco. We were using a Lucent WaveLAN silver card, and already had the driver installed. Surf and Sip network is available free at this point in time, and anyone can use it. We did face problems configuring the card to the network. We created a new network profile and set the network identification (ESSID) to SurfandSip. The card was still not recognizing the network. We tried again with encryption on, but still no luck. We called up customer support at SurfandSip, and had a 10-minute 87

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

conversation with them. There was some problem in the way the ESS-ID was spelled, and that was causing the problem. After that was set right, we opened the browser and we could surf the Internet freely. Surf and Sip does not have any username and password authentication set up yet. In an attempt to gauge the range of the access point, we walked outside the café with the laptop in hand. Transfer rate went down dramatically as we moved away from the café and we completely lost the link after walking about 100-125 feet from the café. Effectively, you could be sitting in your car on the road outside the café and browsing at high speed. We even have pictures doing exactly that ☺.

18.2.2 AirWave
AirWave focuses most on cafes and restaurants and has been pushing its service very aggressively in the San Francisco bay area (Please see interview with the company on page X). The first card we got was a WaveLAN card borrowed from the teaching team. The driver software was downloaded successfully from a website but despite multiple tries could not be installed on the computer, rendering the card useless. The apparent fault seemed to be the inability of the software to launch the install shield wizard ont he specific computer although other software packages were still successful in achieving the same. Hence, a new card was bought and this time installation took only five minutes. This purchase came with a subscription to AirWave which operates in several locations on the Stanford campus like the bookstore and other cafes and restaurants in Palo Alto. We first used the service at The Prolific Oven in Palo Alto. Being a nice day, we set ourselves up on the tables right outside the front entrance but got a rude awakening because there was only a very weak or no signal at all. After inquiring inside, we found that the base station was placed behind a think wooden counter and hence its effective range had been limited to 25 feet! Hence, service within the small cafe was fine but outside dwindled to almost nothing especially if not immediately in front of the shop. This limitation as to the falliability of the access point range was very revealing and surprised us with how inaccurate theoretical ranges can be. Hence we have built this uncertainty not only into our model but from this point onwards decided that site survey and professionals should always do installation only. This decision is reflected in our cost structure for setting up public, home and commercial share zones. One other striking thing was the indifference of some businesses, particularly restaurants, to the fact that they had an access point. This was perhaps because of the fact that AirWave is currently bearing all the cost of these access points. This situation of no cost to location provider can lead to a moral hazard. For example at Gordon Biersch (a bar/restaurant), after much effort we could find only one person who actually knew that the restaurant was even participating in such a thing. 88

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

We asked as to the location of the access point and found out that it was placed in the management offices upstairs (customer presence being only downstairs!). So there seemed to be some disparity as to whom this access point was supposed to serve. Although it makes some sense for restaurant owners to not be too enthusiastic about access points as they have less of a need to allow customers to linger over as opposed to cafe owners. This again strengthened our view that businesses should share some running cost of the access point and should not get a free ride. This would align possible interests in the success of the venture and would be a test for whether the business was willing to put some effort into the operation. Again it is clear that the access point should be placed according to the decision of Share Point’s installation crew. The last place we tried the service was the Stanford bookstore. Since the bookstore has its cafe in a big open space within the building, directly in line of sight of the central access point location, the service was very good at the cafe and almost pervasive throughout the building. Even at over a 50 feet distance, the signal was strong and service was good. But once more to our dismay we found that the signal did not pervade outside the physical structure of the bookstore, dashing our hopes for potential mobile users on the street rather than protable ones who would come in. Login for AirWave was very similar to that for Telia HomeRun although the browser had to be launched manually, after which login directly proceeded to a location specific AirWave web page. There were no complaints about the service or its quality although the Telnet port was blocked, perhaps for valid security reasons.

18.3 Conclusions
As soon as you have tried a few different networks, you’ll get used to how and where to change the network ID for your WLAN card. This, and what address to use for the login page seem to be the only main difference when you’re starting to use a new network. It’s very easy to use, but the usefulness might still have to be proven in some locations. Some locations demand small and convenient devices in order to be able to access the information you want in an easy way. Carrying around a laptop might not always be an option. It’s also hard to read from a laptop screen outdoors. On all locations we experienced high speeds once we got everything installed. People have experienced various problems with wireless Internet access according to our enduser survey. Listed below are the most common ones: Type of problem Technical problems in getting it up and running Low speeds Frequently disconnected Other problems, including bad range and coverage, VPN-client problems Sweden 61% 17% 30% 22% USA 53% 45% 26% 16% 89

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

19 APPENDIX - UNIVERSITY PROJECTS
This section gives a brief overview of some of the university research projects that we have come in contact with while working on this report. Some of them have been going on for a couple of years; others are shorter semester based projects. We have covered projects from Stanford, Berkley, MIT, and KTH.

19.1 Stanford University
There are a number of projects going on at Stanford University. Since three of our team members are from Stanford we’ve had the great advantage of talking to professors and students.

19.1.1 MosquitoNet Research group
The MosquitoNet group (http://mosquitonet.stanford.edu), led by Assistant Professor Mary Baker, claims to be the Mobile Computing Group at Stanford and is part of the Computer Systems Laboratory (For more details on this group’s view of the mobile/portable computing space, please see the interview with Mary Baker). Some of the applications developed by this group are extremely innovative and could form the basis of immediate implementation solutions in the Share Point network. For example, Secure Public Network Access (SPINACH) project has developed an access control that allows only authorized users to access network ports or wireless networks that are available in public areas. This SPINACH solution is already being deployed successfully in the commercial domain. Very briefly, the way it works is that the SPINACH system establishes a "prison wall", controlling the flow of packets between hosts on the Public subnet and hosts outside the Public subnet. As users within the Public subnet authenticate themselves and thus activate network access for their hosts, SPINACH maintains an audit trail. This way, users can be held accountable for any malicious traffic they generate on the network; a key requirement to keep undesirable use of the Share Point network under check. Other potentially useful projects include dynamic bandwidth measurement tools for mobile devices, practical frameworks for personal online identities, Mobile IP for Linux, a Mobile People Architecture that places the person rather than the devices at the endpoints of communication sessions and finally a metadata service that would help people track different versions of files on multiple devices.

19.1.1.1 The Mobile People Architecture
The main goal of the Mobile People Architecture is to put the person, rather than the devices that the person uses, at the endpoints of a communication session. This architecture introduces the concept of routing between people. To that effect, the project define the Personal Proxy, which has a dual role: as a Tracking Agent, the proxy maintains the list of devices or applications through which a person is currently accessible; as a Dispatcher, the proxy directs communications and uses Application Drivers to massage communication bits into a format that the recipient can 90

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

see immediately. It does all this while protecting the location privacy of the recipient from the message sender and allowing the easy integration of new application protocols.

19.1.1.2 Bandwidth Measurement
Mobile hosts and the applications running on them must be able to adapt dynamically to changing network conditions. To do so sensibly, however, they must be able to determine what those conditions are. In particular, they must know how much bottleneck link bandwidth they have. In addition, measuring the bottleneck link bandwidth is important for understanding the performance of many Internet applications. The nettimer bottleneck link bandwidth measurement tool and the libdpcap distributed packet capture library allow fast, bi-directional, passive measurement. In most cases, nettimer can converge to within 10% of the nominal bottleneck link bandwidth after less than 10K of data has been transferred.

19.1.1.3 Secure Public Network Access: SPINACH
Mobile hosts must be able to find some point of connectivity when they move around. This implies that there must be networks, either wired or wireless, that will allow mobile hosts to visit them. However, most organizations only want known visitors to have access to their networks. The SPINACH project (Secure Public INternet ACcess Handler) provides access control that allows authorized users to connect their laptops to your network. In this way, it is possible to leave network ports or wireless networks available in public areas and know that only desired visitors will be allowed to access your network.

19.1.1.4 Mobile IP for Linux
To switch seamlessly between the radio and Ethernet interfaces (and eventually other device interfaces as well), the group has implemented a mobile IP protocol. This protocol allows hosts corresponding with a mobile host to continue to use its home IP address even if the mobile host switches to a network interface with a different address. We implement the mobile host and home agent portions of the protocol, but not the foreign agent. However, interoperability testing has been successful with a variety of other foreign agents and mobile hosts. Our implementation has the added feature that it is possible for mobile hosts to switch dynamically between triangular routing, bi-directional tunneling and regular IP (no mobility support). As mobile hosts move around or network conditions change, the appropriate routing choice may change, and our implementation makes this possible even on a packet-by-packet basis.

19.1.1.5 IdentiScape
IdentiScape is an effort to create a practical framework for representative personal online identities. This research is motivated by the rapid increase of diverse ways in which people can be

91

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

identified on modern communications networks, such as the telephone networks (landline and cellular), and Internet communications (instant messaging, electronic mail, etc.)

19.1.1.6 Roma
The Roma personal metadata service allows people to carry with them a repository of information about their personal files, and helps track different versions of files on multiple devices.

19.1.2 Wireless Systems Laboratory
The Wireless Systems Laboratory (http://systems.stanford.edu/) headed by Assistant Professor Andrea Goldsmith investigates broad areas of wireless systems design to meet emerging technical challenges (for more details on this group’s view of the mobile/portable computing space, please see the interview with Andrea Goldsmith). Current projects are somewhat removed from the scope of our study but include research on a network radio communications infrastructure for automated highway systems and for unmanned autonomous vehicles, ad-hoc networking, wireless video, software radios, and energy-constrained sensor networks. Software Radio is a field that will gain more importance for Share Point particularly when future integration with other wireless technologies happens. Then there will be a need for hardware that can operate in various wireless environments without too much space and weight overhead and software radio technologies will be necessary for that functionality.

19.1.3 SWIG – Software Infrastructure Group
The efforts of SWIG, (http://swig.stanford.edu/) headed by Professor Armando Fox are based on the principle of infrastructure-centric software design: move intelligence from endpoints into the supporting infrastructure. By simplifying the endpoints and concentrating high-level functionality in a logically centralized infrastructure it becomes easier to provide high quality, dependable services. The group is also experimenting with radical new ways of building infrastructures that rely on existing components and services. From hooking COTS data producers and consumers into these infrastructures to building platforms for composing autonomous services without explicit cooperation from the services themselves, this group is at the cutting edge of provisioning that should ideally be implemented in the second generation of the Share Point models. Some of these SWIG projects might prove to be very valuable in the future development of Share Point. The Appliance Data Services group is attempting to “infrastructure enable” mobile input devices like cameras, scanners and microphones so that they will be able to generate outputs that could be put directly to the web. This could clearly be a valuable application that could help create other services. Also, the Interactive Workspace project is developing software that would allow mobile device users to seamlessly enter a room and interact with the applications being used and/or to control various parts of the room hardware without much hindrance.

92

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

19.2 University of California – Berkeley
Berkeley's academic departments consistently rank among the top five in the US and is renowned for the size and quality laboratories as well as the scope of its research and publications.

19.2.1 Daedalus/BARWAN (Bay Area Research Wireless Access Network)
The Daedalus project at Berkeley (http://daedalus.cs.berkeley.edu/), headed by Professors Eric Brewer and Randy Katz did some of the most pioneering work in wireless networking and mobile computing in the late 90s. Admittedly, the pace of development seems to have slowed as many of the core members have moved on to other groups and locations. The goal was to combine intelligent, adaptive applications with smart networking software that could multiplex connections over a wide variety of different networking technologies. Some of the most interesting research with respect to our project involves wireless overlay technologies, handoff protocols and network and link managers. While the many technical and software solutions are relevant to understand, none of them are directly required for our implementations as many have been improved upon already.

19.3 MIT
MIT, Massachusetts Institute of Technology is known for its research centers, laboratories, and programs. There is a lot of activity going on here, and we have by no means covered everything that might be of interest to our study.

19.3.1 Oxygen
The Oxygen project (http://oxygen.lcs.mit.edu/), is an extremely ambitious and large-scale effort with a wide mandate: “enabling people to accomplish more with less work. Bringing abundant computation and communication, as pervasive and free as air, naturally into people’s lives”. The following research areas are currently being covered:

User Technologies Interaction Technologies System Technologies Points of Interaction

Knowledge Access Spoken Software for Change Intelligent Spaces

Automation Visual

Collaboration

Network Mobile Devices

While a lot of the research is very interesting and innovative, it is very forward looking and hence does not always apply to our solutions. However, it must be admitted that reading about the different projects greatly enhanced our understanding of the mobile computing space and the 93

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

technical challenges involved. This is largely a factor of Oxygen’s primary orientation being towards the user experience.

The one project that caught our attention was SpectrumWare. This project is developing new software solutions to wireless communication and distributed signal processing. This allows implementing virtual radios that directly sample wide bands of down converted RF spectrum, and process these samples in application software. The elimination of dedicated hardware by software in this way promises to increase flexibility in future wireless communication systems.

19.4 Cambridge University 19.4.1 HAN (Home Area Networking) group
Since a large fraction of our business model emphasis has been on increasing usage through promotion in the home, we also looked at the HAN (Home Area Networking) group within the Computer Laboratory at Cambridge. It is led by Dr. David Greaves (CTO of Virata) and ambitiously aims to achieve multimedia-capable, auto-configuring, self-organizing, highly reliable, scalable, extensible, secure and easy to use home area networks. Most of their work was tangential to our interests, focusing on appliances and networking issues (all layers) which a give a glimpse into the home of the future. However, we learnt a lot about different wireless LAN technologies from the HAN group and are particularly grateful to Research Student Umar Saif for his constant advice and support in our research activities.

19.5 KTH
The Royal Institute of Technology is the largest of the Swedish technical universities. It provides one-third of Sweden's capacity for engineering studies and technical research at post-secondary level. KTH is very active in the European educational and research programs and also co-operate extensively with well-known universities and research centers in and outside Europe.

19.5.1 StockholmOpen.net
The networks that comprise the Internet are often interconnected in Internet eXchange points (IX) where several network operators exchange traffic. As traffic increases the IX tends to become the bottleneck of the Internet and the demand for more IXs increases. The Swedish ICT Commission (IT-Kommissionen) has announced a vision of one IX per every 30,000 households in Sweden. Thus, the number of IXs will most likely grow in the near future. Most consumer networks are today connected to one ISP (Internet Service Provider) only. This involves a commitment to the ISP for a long time and the direct price competition among ISPs is in practice eliminated. An operator independent access network, where several ISPs are connected can destruct this network dependent monopoly. In this way, the end-users can choose ISP on a much shorter time basis.

94

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

The Kista IP is intended to become an operator independent network, where the students can choose among a number of ISPs53. KTH and the City of Stockholm cooperate on providing an open communication environment based on this idea in the Greater Stockholm Area. Svenska Bostäder (SB), a housing company, Stokab, a company providing dark fiber in the Greater Stockholm area, both owned by the City of Stockholm, and KTH are developing an innovative city network in the Kista area with the following components: ! ! ! Student dormitories at Kista IP are equipped with broadband access via KistaIP.net An Internet exchange (KistaIX), for operators as well as user groups is being established. An open wireless access network, StockholmOpen.net, and a Kista-wide broadband access network are in the planning phase.

These test beds provide great opportunities for students and others to get hands-on experiences in networking, test new innovative services and learn about new business opportunities. KTH IT has received additional funding and support from Stokab to expand the StockholmOpen.Net project. The IT department is now in discussions with a number of interested parties. Skärgårdsstiftelsen, (the Swedish archipelago foundation), want to connect the Stockholm archipelago. Kulturhuset, (Culture house), are interested in providing free wireless access in their building (they are however a bit worried about possible profit loss of their Internet Café). Stokab has put up an access point covering Kungsträdgården in Stockholm. Kista Science Park, the City Hall, and additional parties are interested in this project. KTH is talking to other Universities in Sweden including among others, Chalmers, Mitthögskolan, and LTH, to extend the reach of this project. The ultimate goal is to transform StockholmOpen.net to SwedenOpen.net and maybe even EU.net.

19.5.1.1 StockholmOpen.net proposed business model
There is no existing business model for Stockholmopen.net today. The hope is to provide wireless access to a number of public services free to all users. Services could include tourist information such as museums, restaurants, maps with public restrooms, police, public transportation etc. Other parties might offer services related to their business. For instance, Kulturhuset might provide information about upcoming events, booking of tickets, chat forums, map over the building, etc. If the user wants to access other pages on the Internet, such as the user’s Hotmail account, news, banking, etc., the user picks her or his ISP and can then access any desired web

53

Kista IP-IX Project http://2g1319.ssvl.kth.se/2001/group11/ 95

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

page. Thus the ISPs make money up streams, but the networks are offered for free by location providers and the city of Stockholm.

19.5.2 Communication Systems Design Projects
This project is part of the Communications Systems Design course. In this course there are a number of other interesting projects focusing on wireless network solutions. Some of these projects will end after this semester; others will live on, either in the sponsoring companies or at KTH by PhD students.

19.5.2.1 The Turncoat Project
The aim is to investigate how to allow mobile phone roaming between two different media, WLAN and GPRS. The phone should sense which media provides the best service and switch when needed - hence the project name "TurnCoat”. The intermedia roaming will be handled in the network layer, so the two protocols that support mobility, Mobile IPv4 and IPv6, will be investigated for this purpose.

19.5.2.2 BLUMTS
The goal for this project is to achieve handover functionality between two Bluetooth neighboring cells and a handover from/to the GPRS network. The deliverable includes technical realization of the two handovers, academical aspects (physical limits), technical and economical aspects of the Interface to existing operators.

19.5.2.3 Flying Freedom
The objective is to provide anonymity, intractability and unlink ability services in wireless networks. The main goal of the project is to integrate freedom (high privacy software) in a FlyingLinux.NET-like wireless network using the open source freedom client from Zero Knowledge Inc. and to study the possibility of running freedom in the so called ``3G'' radio networks. The first objective is to integrate freedom in the KTH/IT University "Yellow" network, an IEEE 802.11b based network. The second is to integrate freedom in a MobileIP network.

19.5.2.4 Hottown
To produce and verify the relevant principles for supporting the delivery of personalized and adaptive services in mobile networks and demonstrate these in a distributed wireless packet network testbed. The assumption is that mobile individuals are in the possession of small (creditcard size) low-cost lightweight, packet-radio- (GPRS, WLAN), multimedia, positioning- (GPS and accelerometer) and Bluetooth/IrDA-enabled communicators with hands free (used as default interface), a small touch-sensitive screen, including a small camera and attachable personal display, all in one neat tiny package (see graphics).

96

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

The deliverables include Personalized and Adaptive Services, Application Architecture and Mobile Device Software Architecture. Ericsson Research contributes with SIP, mobile agent technology and if applicable access to a wireless packet network test bed.

19.5.2.5 Kista IP-IX
The Internet consists of a large number of interconnected networks. The networks are often interconnected in Internet eXchange points (IX) where several network operators exchange traffic. As traffic increases the IX tends to become the bottleneck of the Internet and the demand for more IXs increases. The Swedish ICT Commission (IT-Kommissionen) has announced a vision of one IX per every 30,000 households in Sweden. Thus, the number of IXs will most likely grow in the near future. Most consumer networks are today connected to one ISP (Internet Service Provider) only. This involves a commitment to the ISP for a long time and the direct price competition among ISPs is in practice eliminated. An operator independent access network, where several ISPs are connected can destruct this network dependent monopoly. In this way, the end-users can choose ISP on a much shorter time basis. The Kista IP is poised to become an operator independent network, where the students can choose among a number of ISPs.

97

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

20 APPENDIX - MATRICES
20.1 Exhibit 1: Wireless LANs and PAN standards comparison matrix 20.2
Use Max Physical rate (Theoretical) Max Physical rate (Actual) Range Bluetooth IrDA HomeRF IEEE 802.11b LAN HomeRF/2 IEEE 802.11a LAN HiperLAN/2

PAN

PTP

LAN

LAN

LAN

(1 Mbps) 400 Kbps (720 Kbps) 10 m

1 Mbps

1.6 Mbps

11 Mbps

10 Mbps

54 Mbps

54 Mbps

1 Mbps

1.6 Mbps

11 Mbps

-

-

-

1m Infrared region In dev (Shipping)

50 m

100 m

50 m

100 m

100 m

Spectrum Frequency selection Status

2.4 GHz

2.4 GHz

2.4 GHz

2.4 GHz

5 GHz

5 GHz

FHSS In dev (Shipping)

FHSS

DSSS

FHSS

OFDM

OFDM

Shipping

Shipping

In dev

In dev

In dev

20.3 Exhibit 2: Major Wireless WANs standards comparison matrix
GSM Europe, USA CDMA one USA, Japan TDMA PDC GPRS CDMA 2000 USA, Japan 144 Kbps 614 Kbps* 2 Mbps** 144 - 384 Kbps CDMA 2000* CDMA 2000** In dev EDGE USA, Europe 384 Kbps 100 Kbps WCDMA Japan, Europe, (USA) 384 Kbps, 2 Mbps

Major markets Max Physical rate (Theoretical) Max Physical rate (Actual)

USA

Japan

Europe

9.6 Kbps

14.4 Kbps

9.6 Kbps

9.6 Kbps

100Kbps

9.6 Kbps

14.4 Kbps

9.6 Kbps

9.6 Kbps

38 - 62 Kbps

384 Kbps

Evolution

GPRS later EDGE Shipping

CDMA 2000

TDMA+ later EDGE Shipping

PDC+ or WCDMA

EDGE

-

-

Status * Second generation

Shipping

Shipping

Shipping

In dev

In dev

** Third generation

98

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

21 BIBLIOGRAPHY
! ! ! IDC Research “Unwiring the Network: Worldwide Wireless LAN Market Forecast Update, 2000-2005”, April 2001. Mobile Internet Magazine, Volume 2 Number 7, April 6 2001. p 9-10 Victor Bahl, Anand Balachandran, Srinivasan Venkatachary, The CHOICE Network: Broadband Wireless Internet Access In Public Places, February 2000 Technical Report MSR-TR-2000-21 Worldwide Wireless LAN Market Forecast and Analysis, 1999-2004, IDC Report #23431 - December 2000. The Pulse. Networking Magazine, November 2000. Den svenska telemarknaden fortsätter att växa. Post och Telestyrelsen. March 9, 2001. Den svenska telemarknaden fortsätter att växa. Post och Telestyrelsen. May 9, 2001. Nokia Operator Wireless LAN Solution Brochure. The Standard, “Peace, Quiet and a T-1 Line”, May 3, 2001.

! ! ! ! ! !

22 WEB RESOURCES
! ! ! ! ! ! ! ! ! ! ! ! http://ieeexplore.ieee.org/lpdocs/epic03/ http://www.homerf.org http://www.Bluetooth.com http://www.etsi.org http://www.ericsson.com http://www.nokia.com http://www.atheros.com/ http://www.wireless-nets.com/whitepaper_interference.htm http://www.oreillynet.com/pub/a/wireless/2001/02/23/wep.html http://www.planetanalog.com/features/rf/OEG20010510S0037 http://www.hpl.hp.com/personal/Jean_Tourrilhes/Linux/Linux.Wireless.std.html http://alpha.fdu.edu/~anandt/

! ! ! ! ! !

http://www.homerun.telia.com http://www.stockholmopen.net http://www.sydkraftbredband.com http://www.powernet.se http://www.wirelessbolaget.com
http://www.abrandnewworld.se

99

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

23 ACKNOWLEDGEMENTS
We wish to thank Ericsson Business Innovation and Telia, especially Mats Segerström, Jasminko Mulahusic and Roland Larsson for sponsoring our learning process, and for their continuous participation and interaction throughout the duration of the project. We are thankful to our coaches, Lena Ramfelt and Magnus Johansson, for their constant support throughout this project. Their participation, encouragement and feedback were invaluable to us. We are grateful to Prof. Bjorn Pehrson of KTH and Prof. Tom Kosnik of Stanford University for their mentorship and time during this project. We would like to thank the following for their time and invaluable insights during our interviews with them: ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Prof. Mary Baker (Dept. of Computer Sciences, Stanford University) Prof. Andrea Goldsmith (Dept. of Electrical Engineering, Stanford University) Prof. Donald Cox (Dept. of Electrical Engineering, Stanford University) Mr. Ross Goldberg (VP Marketing, Airwave) Mr. Rick Ehrlinspiel (CEO and President, Surf and Sip) Mr. John Pardun (National Accounts Manager, MobileStar Corporation) Mr. Manav Khurana (Assistant Product Manager, Symbol Technologies) Mr. Brewster Kahle (Founder, SF LAN) Mr. Stephen Oskoui (Student, Stanford University) Mr. Hosain Rahman, Chief Strategy Officer, Aliph Communications, San Francisco. Mr. Umar Saif, Reasearch Student, Home Area Networks project, Computer Laboratory, Cambridge University. Mr. Craig Barrett, Chief Operating Officer, ArrayComm, Santa Clara. Mr. Khurram P. Sheikh, Chief Technical Advisor, Sprint Broadband Services. Mr. Hussein Kanji, Stanford Alumnus. Prof. Björn Pehrson (Dept. of Computer Sciences, KTH) Mr. Alberto Escudero (PhD student KTH) Mr. Jonas Willén (PhD student KTH) Mr. Fredrik Lilieblad (PhD student KTH) Mr. Rune Gunnarsson (Founder, PowerNet) Mr. Bo Gråberg and Mr. Pontus Nord from Wirelessbolaget Mr. Joakim Nilsson and Mr. Miguel dos Santos from Telia HomeRun Dr. Bo Magnusson from A Brand New World Mr. Anders Persson from Sydkraft

100

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

24 PERSONAL THOUGHTS OF THE TEAM MEMBERS
24.1 David Alvén
A great project which I've really enjoyed working on; also being a part of the BATLAN team has been a great experience. I've learned a lot and I feel for once that a school project has generated something useful. It has been very interesting to work on a complete project with both the technical aspects as well as the financial aspects.

24.2 Resmi Arjunanpillai
It is definitely one of the best projects that I have been part of - both in terms of the project objectives and the project team. I learned a great deal about the wireless LAN space and about working effectively in teams. I can honestly state that most of the personal objectives that I had in taking up this project have been met. The trip to Sweden was an added bonus. Three cheers to the BATLAN team for an awesome experience.

24.3 Reza Farhang
This project / course has been the most interesting and instructive course during my time at KTH. I have learned more about working in a project team then I could ever imagine, and hopefully it will be a good help in my professional career, at least during my thesis project. I would like to thank our sponsors and coaches for their involvement.

24.4 Sachin Kansal
I have been very impressed by the kind of interest the sponsors have expressed throughout the duration of the project. Frankly, I have never been a part of a project where the sponsors were deeply involved at each stage. They helped the team learn and grow by encouraging us and prodding us on. I believe that we have been able to live up to the expectations and that our proposed solutions will add value to Telia and Ericsson going forward. It’s been quite a ride!!!

24.5 Nauman Khan
I think the Share Point idea is a great business opportunity. It positions Telia ISP very formidably, considering nobody has thought of building up a base of wireless LAN users from the home and very few are even capable of doing it. Don't wait for the killer applications; they will come (Access is the killer app!). At these price points, people can just choose to surf. Telia ISP should not let this one go.

101

Analysis of a 3G Alternative – Final Report Communication Systems Design 2G1319

24.6 Ulrika Leufvén
I'm very glad that I could be a part of this project. In addition to deep knowledge in the wireless market space, it has taught me a lot of things about myself - things that will be very useful to know now that I am to start looking for a job. Working together with all the team members has been a lot of fun and I realize that I need to share my energy and thoughts with others. I also discovered how working in a project affects your personal life. In the beginning, it is really hard to dedicate a lot of time because you're not quite sure what you're doing. In the end, you sleep about 5 hours at night, don’t have much time for friends and other activities and spend all your time and thoughts on the project. I'm sure that if you're more used to working on projects you can plan your time better. Nonetheless, I think it is a quite common phenomenon.

102


								
To top