Albert, Steve, and J.R.,
Albert, Single-pair 2 megs/s HDSL (the animal the press release and ETSI were referring to) did not exist in early 1995. The ETSI announcement was last week. So the fact that you found nothing on it in a 1995 doc doesn't tell us much. As we both know, the 1996 ETSI press releases were not available on their site, and standards monitoring info was not available to nonregister users. If you really feel there is something amiss, why not e-mail ETSI in France?
Steve, Are you thinking of cross-talk here (interference from adjacent wires)? You're right that there's nothing extraordinary about that (and the lastest generation CAPs have excellent cross-talk protection with their trellis coding). The experts at Telchoice used the phrase "extraordinary circumstances" when referring to this loss of a part of the copperwire spectrum.
J.R., The only place I hear this noise to signal ratio criticism of CAP is on the Amati threads. The trialing telcos, the Bell Atlantic customers who were thrilled over their VoD with Westell Flexcaps last February (in northern VIrginia), the experts at Telechoice, industry journalists in EE Times, etc., no one else seems to have noticed this. The difference, as all the reports of the trials say (according to Danny Briere's posts, before he was hounded from the Amati threads for his heresy), is not noticeable in field trials.
Two points on prices: (1) the chipset is only a fraction of the total cost--in the current DMT prices, the chips make up no more than $300; and (2) as you say, there is no free lunch, and spectral decomposition a la DMT costs plenty. My feeling is that this superrefined decomposition is not needed. If you can offer some reasons why it is (and under what line conditions), I'm open to any well-reasoned and substantiated ideas.
Dave |
