| UNIVERSITY OF PENNSYLVANIA - AFRICAN STUDIES CENTER |
The COOK Report on Internet -> NREN announces the publication of a Special Report: Wireless as an Internet On-Ramp and Local Loop By Pass Technology. This 22,000 word study (available now and thoroughly documented via cites from the relevant published literature of the past 6 months) will be serialized in the September through December issues.
We find that, by the summer of 1995, wireless bridges into Internet service providers and wireless providers themselves can be expected to be accessible in the largest urban areas. These services, based on unlicensed spread spectrum technology, are also feasible and economic for use in rural areas. Of course they won't be "rolled out" *there* unless a group of users puts together the necessary funds to pay for them. Still the bottom line is that reasonably priced wireless internet solutions are available.
Introduction - We survey wireless technologies from a data rather than voice perspective. Since most reasonably priced wireless solutions depend on spread spectrum, we immediately look at the question of whether a spectrum shortage exists and whether it will inhibit the development of this technology. We compare and contrast Gilder's views on Steinbrecher Digital radios, TDMA and CDMA, and the politics of spectrum allocation with those of Andy Seybold in his book Using Wireless Communications in Business. Gilder sees spread spectrum as an absolutely revolutionary technology that is eliminating spectrum shortage. Seybold sees it as an important technology but one unlikely to grow far beyond its current applications because there is a severe spectrum shortage and the needs of the big players will keep spread spectrum tied up within the bounds they establish. We conclude that the truth lies somewhere in between the two positions - why? Because although technology advances are making amazing improvements in spectrum use possible, the economic drivers behind business and regulation act with some consistency to inhibit progress.
We conclude that while spread spectrum systems certainly do not offer a risk-free path to any communications utopia, they are viable in the current 902 to 928 MHz bandwidth and can be used effectively in the 2 and 5 GHz ranges where they are also authorized.
As part of our research on the viability of spread spectrum we looked at the 1981 to 1985 FCC rule making process to understand why the FCC allowed spread spectrum in the first place. We were assured by a knowledgeable Washington observer that the FCC never intended to open the gates much beyond the point reached by the final 1985 rule making. Nevertheless, the May 1984 FCC notice shows that after 3 years of technical comment and study the Commission was prepared to allow *unlimited use* of the technology in bandwidth above 70 Mega-hertz, and *unlimited power* in the ISM bands so long as such use did not interfere with other licensed use in those bands. We were told that what was finally approved a year later (one watt in the ISM bands only) was an evisceration of the technology by key large corporations and the security agencies which complained that they'd never be able to monitor the bad guys under the conditions of the 1984 proposed rule making.
Part One of our study looks at mixed uses of voice and data. It focuses mainly on CDPD which appears not well-positioned for the long haul. It examines digital cellular and PCS to see where these technologies fit in in view of the explosion of money at the first set of frequency auctions - which it sees as *not in the public interest* and likely to kill the viability of PCS.
Part Two looks at Wireless Two Way Data Services and add ons. We explain RAM Mobile Data and ARDIS and the features differentiating their services. We look at RadioMail and how both RAM and ARDIS have used it. Finally we look at emerging new services - WYND as a RadioMail competitor and Nationwide Wireless and Nextel.
Part Three examines Metricom's Micro-Cellular Data Network Service in considerable depth. Metricom is a data network service that with over $100 million in the bank is poised to roll out a *low cost* intelligent mesh network service using $700 radios that can be hung from light poles and the sides of buildings. As the network grows in number of users and traffic, the infrastructure is expended by throwing more $700 radios in place. The low cost is made possible by use of spread spectrum technology and by not building in the capability for rapid user hand offs between cells. You can WALK from cell to cell while online, but not drive.
Metricom's first commercial network, in operation in the Silicon Valley area since June 28 1994, is delivering 38.4 kilobaud speed flat rate and unmeasured service for $29.95 a month. This is twice RAM's best speed and an unmetered rate far less than anyone else in the industry where metered rates predominate.
We publish a telephone interview we conducted with Metricom President Bob Dilworth. In that interview we clarified Metricom's offering of Internet PPP access for $15 additional dollars a month. Metricom does not give users a mailbox on the net. What it does is deliver them to a gateway to BARRnet. Users have to pay regular rates for BARRnet services IN ADDITION to the Metricom charges.
Part 4 Tetherless Access: we interview TAL President Dewayne Hendricks. TAL is more directly oriented to the provision of Internet service than Metricom. TAL has a spread spectrum wireless modem/router at $3500. These devices can easily communicate over line-of-sight distances of as much as 20 miles. We also interview Richard Lowenburg the Director of the Infozone project in Telluride Colorado where TAL will turn on a network between now and Labor Day.
Conclusion: the degree of success in Metrocom's roll out over the next year will be critical in determining the extent to which, by next summer, wireless can serve as effective low cost Internet access and telco by-pass. We note that markets in rural areas could be effectively served by a combination of a Metricom radio cluster in a small town and Tetherless daisy chain of radios from a telco pop to that town - *IF* Metricom decides to provide a low cost gateway to its base station. (Metricom radios do *NOT* daisy chain well.)
Our report is available now at an invoiced price of $250. Current subscribers to the COOK Report May purchase the entire report NOW for $150. We expect that the Metricom section of the report will not be published before December.