Frank, my initial impression of the AT&T / TCI engagement is that it could result in an exponentially greater financial disappointment to AT&T than their failed foray into computer hardware with NCR. AT&T at Basking Ridge and TCI in Denver are as diametrically opposed (in a corporate cultural way) as any two entities in the USA. On one hand, AT&T still has the legacy of things like the 30 volume "Bell System Practices", a set of procedures so comprehensive that today, fifteen years after their divestiture from AT&T, the five (soon to be four) RBOCs still conduct their local exchange telephone business in a remarkably similar manner. On the other hand, there is TCI. TCI is hodge podge of cable systems acquired from smaller cable operators over a period of a few decades. Most of these systems have technology and methods of operation that are quite different from each other, with very little in the way of standard practices....in other words, the exact opposite of how AT&T likes to organize its entities.
Over half of the TCI systems have yet to be upgraded to hybrid fiber coax with upstream communications bandwidth - this pretty well explains why wireless local loop is part of the last mile equation in an AT&T / TCI network proposal. Even those TCI systems with both HFC and upstream bandwidth will struggle to make the coaxial return path clean enough to use for IP based voice and data services. TCI tried and discontinued voice services over a HFC network in Fremont, CA. Part of the problem was economics, and part of the problem was technical with return path noise ingress. Even more troublesome for AT&T than the overall technical level of TCI networks may turn out to be the less than stellar level of service and customer relations in many areas. It won't be an easy sell for AT&T/TCI local telephone service to customers who are not very satisfied with their TCI cable service.
Of course, I cannot leave the subject of voice over IP via cable or WLL without pondering the power issue (one of my favorite subjects). Cable systems are still very vulnerable to service interruption during a power outage. Unlike copper pair telephone networks that are powered from the central office (or from remote digital loop carrier cabinets with at least 8 hours battery backup), most cable systems, including those with HFC,have limited battery reserve in case of power failures. Will the VoIP customer premises equipment be powered by the customer? If via WLL or cable modem, I think so.
The AT&T / TCI deal provides the RBOCs with more big ammunition to gain entry into long distance within the next year or so. Once given the regulatory "green light", long distance will be available to a large portion of RBOC customers within a couple of months. Much of the planning is already complete, and the separate long distance subsidiaries are already established and pretty well staffed. The point is, it would take a RBOC a lot less time to get into long distance and tear off a sizable chunk of AT&T long distance revenues, than it would for AT&T to get a sizable portion of local RBOC customers via TCI. Use SNET (Southern New England Telecomm.) for an example in Conn. Once given permission to offer long distance, they attracted around 1/3 of their local telephone customers to sign up with SNET long distance service in less than a year. (Now you know why SBC really wants to buy SNET!.
Off topic, Frank, you asked about cross linking some of these posts with the last mile thread. Fine with me.
Ray. |
