Tim, here's the same article from the same reporter, but it seems to have slightly different information. Regards, Mark
AT&T's Faith In Cable Justified?
By Fred Dawson, Contributing Editor
July 12, 1999 9:25 AM ET
AT&T is preparing to field test a new approach to cable network design that could go a long way toward vindicating its high-stakes gamble on the medium's broadband service potential.
With $120 billion committed to cable acquisitions, AT&T has been hard-pressed to justify future cost projections that envision drastic reductions from the current cost of delivering voice, data and advanced video services over Hybrid Fiber-Coax (HFC) networks. Now, with a new approach to distributing signals made possible by the new network design, there's a chance the company could drive down the costs of end-user terminals and other gear to levels that meet or beat those projections, officials said.
AT&T Broadband and Internet Services plans to run the trial of the new network design in Salt Lake City over 520 miles of cable plant passing 66,000 households starting in October, said Xiaolin Lu, district manager at AT&T Labs.
The first phase, to be completed by the end of the year, involves extension of fiber deep enough into the coaxial distribution plant to eliminate the need for amplification of signals between what are called "mini fiber nodes" and the customer. This increases the usable bandwidth available over the coax from the current ceiling of about 750 megahertz to at least 1 gigahertz - a gain of more than 30 percent. The approach reduces the homes served by the coax to between 50 and 100, vs. the 600 or so now served from fiber termination points, Lu said.
If this phase of the test bears out the cost and performance assumptions of the new design, the company will move to the second phase in which a new approach to transmitting voice and data services could produce big savings at the end-points.
"When we move into phase two, we have a potential savings in user terminals," Lu said. "Not only do power consumption and costs go down, but performance goes up."
This is because the company would use the traditional Time Division Multiplexing (TDM) mode common to data transport applications such as 10BaseT or 100BaseT Ethernet for new classes of Internet Protocol voice and data services, which would be added at the frequency tier made available by the expanded bandwidth of the coaxial links, Lu said. In other words, the company would be able to use off-the-shelf components rather than relying on the expensive and complex chips used today at the headend and in user terminals to perform the media access control, modulation, conditional access and quality-of-service functions that make IP communications possible over cable.
AT&T BIS engineers estimate the cost of upgrading to what they call the Lightwire design beyond the costs of the traditional two-way HFC architecture will be roughly $40 per home passed. The company expects operational savings of about $11 per home per year because of lower power consumption and reduced maintenance gained by eliminating 61 percent of the active components in the coaxial plant.
On this basis alone, the topology could prove out for upgrades of older plants that haven't been upgraded to state-of-the-art HFC designs, such as many of the systems operated by AT&T, said Marty Davidson, district manager for product realization at BIS. But, if phase two demonstrates the viability of the TDM-based approach for advanced services, the cost benefits could extend to a much broader plant base sooner than later, he added.
While cable operators are closely watching what AT&T does, many engineers at other companies remain skeptical that use of so much fiber and new types of electronics will cost as little as AT&T anticipates. They expect the falling prices of customer premise equipment built to the new cable data modem standard to eliminate any cost advantage AT&T's approach might have in that arena.
zdnet.com |
