ADSL hitches and glitches
abound
Only a select few users are likely to enjoy asymmetric digital subscriber line service in
the next two years. At least that was the prevailing wisdom at the second ADSL
Forum Summit, a gathering of developers, carriers, and analysts held in Boston last
month. Enthusiasm about ADSL's potential, buoyed by news of ecstatic trial users, was
tempered by the economic and technological realities of real-world, large-scale
deployment issues that will not be solved overnight.
Before telcos begin to offer services, they want to be sure they will get a substantial
return on their maintenance and deployment investments. But there are also
technological reasons for delaying.
Nigel Cole, technical committee chair for the ADSL Forum, acknowledged that
carriers still must remove load coils and severe bridge taps from some lines in order to
effectively deploy ADSL. Load coils block ADSL's high frequencies, and bridge taps
reduce the signal and produce reflections on the line.
Cole also said that there are crosstalk problems with some older lines.
Perhaps most ominously, Cole said that abnormal customer premise wiring with high
interference levels has caused problems in trials. In many cases, it appears that internal
wiring in certain homes will have to be replaced in order to deliver ADSL.
"It's difficult to conclude now how much we can use the internal lines [in homes]," said
Cole.
Although the basic technology works, the telcos and alternative providers must also
resolve management, provisioning, billing, and deployment issues between regional Bell
operating companies, competitive access providers, and ISPs (Internet service
providers).
Bobbi Murphy, an analyst at Dataquest in San Jose, Calif., warned against overhyping
ADSL by drawing a comparison to the hype-rich but disappointing rollout of ATM.
"You have to be careful not to build up expectations that the technology might not be
able to live up to for quite a while," Murphy said.
While Cole envisioned 1999 as the breakthrough year for ADSL, Murphy predicted
the technology would be in an early-adopter phase for the rest of this decade, with
widespread deployment not beginning until 2001.--Joe Paone |
