Why this column?
Wireless Internet Access, in its many forms, is increasingly of interest to Internet Service Providers. Many wireless digital products are intended for markets such as wireless lands and mobile e-mail. This column will try to cover wireless data developments as they relate to Boardwatch's audience of Internet Service Providers.
Is it practical to use wireless digital technologies to provide Internet access? Yes! When you consider that wireless digital technologies are being driven forward by the same forces as microprocessor technology, that answer is unsurprising. It's a certainty that most ISP's will become involved in wireless digital in the next several years as it becomes more capable, easier to implement, and less expensive. Your customers, and your competition, will demand that you do so.
An excellent illustration of wireless digital's promise was AT&T's announcement on 2/25/97 that AT&T Wireless and AT&T Labs have developed a practical wireless replacement for the conventional telephone distribution system based on copper cabling. AT&T developed this system, whose development name was Project Angel, as a way to provide local telephone service without having to build its own cable infrastructure to the home, or pay "rent" to established telsons for the use of their lines.
There is much to admire about AT&T's fixed wireless telephone system. One of its most surprising aspects was that it was developed in relative secrecy in Redmond, WA. AT&T apparently developed the system, which combines several digital modulation techniques- Time Division, Frequency Division, and Space Division to be able to effectively service thousands of users in the 10 MHz of spectrum that AT&T obtained. The secrecy enabled AT&T to bid relatively reasonable prices in the recent spectrum auctions, obtaining allocations sufficient to cover 93% of the US population.
AT&T's system relies on base stations that could service between 1000 and 2000 fixed wireless telephone users. The base stations will often be the same, existing base stations that service AT&T's mobile cellular telephone customers. AT&T has reserved space for the fixed wireless telephone equipment at its base stations for the past two years, and a relatively small 30" semi-circular antenna needs to be added to the antenna systems. The base stations are then connected to conventional telephone switching centers via fiber optic lines.
At the home, a "tranceiver the size of a medium-shaped pizza box" (18" square) is mounted on the side or rear of the home, and the home's telephone wiring is routed to the tranceiver. Since the home's phones would no longer be powered by the central office, a battery backup unit "about half the size of a desktop computer" can be installed.
AT&T's fixed wireless telephone system provides at least 2 telephone "lines", and Internet service at a minimum of 128 KbPS. Left unsaid is how a computer in the home is connected to the tranceiver- additional 10baseT wiring between the tranceiver and computer is probably required. Transmissions between the base station and the home transceiver are encrypted. The system could include the ability to use the same wireless phone as a cordless phone (usage charged at "home" rates) or a mobile phone (usage charged at "mobile" rates). Also left unsaid is the probability that the Internet services will be "always connected", since when there is no traffic, no "resources" are being consumed (unlike with a conventional telephone line and modem connection which ties up a home to central office to SIP circuit even if no data is being transferred at a given moment). There is also the possibility of adding a higher speed Internet connection (downlink only) by using DirecPC, the satellite-based Internet service from Hughes Electronics. DirecPC can transfer data from the Internet to the user at 6 MbPS, but still requires a conventional Internet connection to transmit data from the user to the Internet. DirecPC technology could also be used in terrestrial systems. AT&T has invested in DirecPC.
Formal testing of the AT&T's fixed wireless telephone system will be conducted in Chicago in 1997, initially with AT&T employees, and later involving the public. Those tests will help determine what the pricing of the system and service will be. Components for the system are already being manufactured at AT&T's Redmond, WA facilities where the system was developed. Future enhancements may include faster Internet access and full-motion videoconferencing.
AT&T's fixed wireless telephone system has enormous potential. The savings in labor and maintenance costs alone could be enough to give AT&T a decisive advantage in pricing local telephone service against telcos using copper cabling. Copper cabling is easily damaged by backhoes, falling trees, water infiltration, and offers only limited bandwidth for Internet connections. Copper cabling must also be maintained by skilled, expensive technicians. AT&T's fixed wireless telephone system could easily be upgraded to incorporate new technologies such as faster Internet service and full motion videoconferencing by downloading new code into non-volatile memory, or at worst, swapping out the tranceiver.
AT&T can even avoid investing in labor to service the "last remaining copper"- that of the telephone wiring installed in the home. It's long established that the homeowner is responsible for telephone wiring past the demarcation point, which in this case would be the tranceiver. AT&T could easily contract out the installation and maintenance of the tranceivers and the home telephone wiring, or the customer could choose their own telephone contractor. (This would finally realize Jack Rickard's long-held dream of the "Rusty Pliers one-man-and-his-truck telephone company".)
The cost savings of AT&T's fixed wireless telephone system would not just be on AT&T's side. The customer would benefit by AT&T being able to offer bundled, lower cost services such as "additional" fixed telephone lines, Internet services, cellular/mobile services, videoconferencing, etc. AT&T's ability to bundle these services (and do so with inherent efficiencies, unlike other bundling arrangements where multiple services are merely combined onto one bill) should generate some compelling cost savings.
Will AT&T actually deploy this system? I, for one, hope so. My home, not too far from Redmond, is subject to all the typical problems of suburban copper telephone cabling- low speeds, service outages during storms (lots of big trees), and general lack of willingness of my current telco to do much about making ISDN or faster connections to the Internet affordable. The AT&T fixed wireless telephone system seems ideally suited for fast growing (profitable) suburban areas, but less well suited for densely populated urban areas.
Sources of info for this story:
Wall Street Journal, 2/26/97, Page B3
Eastside Journal (Bellevue, WA), 2/26/97, Page B1 - B2
http://www.att.com/press/0297/970225.pca.html
Post-publication Notes
This article originally appeared in Boardwatch Magazine, April, 1997. The contents of Boardwatch Magazine is no longer available online; the current boardwatch.com site has no relation to Boardwatch Magazine.
As with all archive postings, the links mentioned in the column above may no longer work, and are included only for historical purposes.
The content above is "as submitted" for publication; only typos and other minor errors will be changed, if at all.
This was the debut piece of my professional writing career (though I'd had one other paid piece of writing more than a decade earlier), and the very first Wireless Data Developments column.
AT&T did actually deploy Project Angel and it ultimately performed "reasonably" well; "reasonably" varies depending on who you talk to. The third major revision of the Project Angel base station system and customer premise equipment worked reliably. Customers liked Project Angel because it allowed them to have flat rate long distance at home and broadband Internet access (initially 512 Kbps, ultimately 1 Mbps) where DSL and cable modems were not deployed. Financially... the costs were never... and according to some analysts, never could have been recovered from Project Angel revenues.
One of the most notable things about Project Angel was that it was one of the very first systems to use Orthogonal Frequency Division Multiplexing (OFDM).
Project Angel ultimately fell victim to the collapse of then-Chairman Michael Armstrong's plan to offer local and long distance telephone service via wireless and cable television, financed by debt and long distrance revenues... which collapsed. To recover shareholder value, Armstrong spun out AT&T Wireless Services and Project Angel was included as part of that spinout.
AT&T Wireless Services quickly shut down Project Angel (it clashed, mightily, with their vision that their business was mobile wireless...) and sold it for a relative pittance to Netro, who planned to rearchitect it for use in overseas markets. Netro pulled that off... barely... but ended up being acquired by SR Telecom. At last report, SR Telecom is enjoying modest success with the "overseas" version of Project Angel.
Steve Stroh
November 15, 2004