AHhaha, your assumptions re: the exponential rising costs of DSL are interesting, but I don't fully understand them. They seem to be a wholesale contradiction to the wisdom of economies of scale and the powers of VLSI, and service integration. Please explain.
The ILECs will loss-lead if they have to, no doubt, in order to eliminate or suppress the competition once they recognize that the competition is real enough to begin eating into their voice and other traditional services. One has to keep in mind here that voice is inextricably attached to loop rentals and in many future instances, long distance, as well. For these reasons alone (and there are many others), this makes voice a pivotal issue to contend with. I therefore would not discount or trivialize its relevance, from this strategic standpoint, alone.
It's going to take a considerable amount of time, if even in the foreseeable future, before the daily telephone calling habits of most personal and business users shift from a POTS model to one that is screen-based or derived from a super-UDP VoIP, such as MPEG.
VoIP uses User Datagram Protocol, or UDP packets, which do not require the establishment of sessions. These are like cockroaches which can find their way through the smallest cracks of available bandwidth, while suffering considerable packet loss at no distinguishable difference to the user. This is due to predictive sampling and correction in the algorithms used. MPEG can't do this as easily, and it doesn't survive well in hostile environments yet.
And just so we don't prematurely fall trap to the notion of the "ubiquity and freeness of bandwidth," let's keep in mind that most [and eventually all] cable modem providers have, or will, begin imposing use policies that will seem ultra restrictive to end users which result in their feeling somewhat cheated, since they will feel that they have been deprived access to the ultimate "window." This is because SPs are either already feeling the bandwidth pinch in their infrastructures, or they can see it coming.
These realities stem from inadequacies in their initial platforms in this space with overcrowded Ethernet-like designs, and doomed-from-the-start bandwidth limitations of the original ADSL model (reserving that very-high-speed DSL, or VDSL still holds out some limited hope for some of the incumbent carriers, but at an even greater cost of entry than ADSL).
This is not to suggest that the DSLs or the wireless guys will be in any better shape. They will be worse off at some point, likely right from the beginning. Rather, all providers with their current architectures will be subjected to onerous bandwidth budgeting routines and route cost allocations, regardless of which popularized model we are speaking about at this time.
@Home has already imposed use policy restrictions;
UMG has done it at the horrification of some of their users;
DirecPC has their fair use policy limits which now see them in court facing a class action suit;
They will all do it. Until their architectures shift to the use of waveguides that we have been discussing in the SR thread, or until they radically reconfigure their current provisions, they will all impose curfews and payload limitations on their users traffic patterns.
By the way, see the latest paper fromTelecommunications Magazine in the SilkRoad thread which speaks about:
"The Role of Optical Internets in the New Public Network" at:
Message 7336986
------------------
MPEG delivers a far superior level of audio quality, hands down. At the same time it is still quite fragile and requires special "session" handling and in some parts of the network it requires forward error correction, among other considerations. Otherwise, it is prone to breaking up, which would result in distracting the user thus defeating its original intent of superior quality. All of these special handling requirements currently eat into the limited processing and bandwidth resources at both the head ends and at end-user locations at this time, and will result in further choking the topologies which separate central from remote.
These problems will go away, they will be a no brainer, when outside goes optical, but for now it's just another burden to this still-evolving and contentious (from a protocol standpoint) and resource-deprived model that we call HFC. This holds true IMO for all of the configurations that I have seen thus far, when you take into account the expectations that users now have w.r.t. m-m, streaming or real time, gaming, personal video conferencing, etc.
All of this having been said, the HFC cable model now stands poised in the forefront to be a winner, even from the standpoint of it being the least onerous of all bad choices, despite some of its obvious and inherent shortcomings and the latent, and now sneak-up-from-behind-demand, that it now faces.
But as always, the real test will lie in its ability to recognize the need for reconfiguring at least the outside portion during its next structural revamp, utilizing optical techniques to the fullest allowed by price performance and other metrics.
In the interim the RBOCs are facing the same challenges, probably more so due to the inferior speeds allowed by DSLs from the start, and they will likely resort (if they have the necessary temerity and solvency of mind and spirit by then) to an extended fiber-to-the-neighborhood/full service area network (FSAN), to counter such a move. The latter comes eerily close to that of fiber to the curb [FTTC] variants of cable designs which were discounted as being too expensive, just a few years ago.
FSAN is a form of passive optical networking, which is the closest large-scale-orchestrated model of standards that I've seen to bringing fiber close enough to the user to where it makes a difference. It does not, however, bring the silica into the user's window. To date, only a handful of communities have received variants of FSAN, and I know of none presently that are underway. Maybe someone can enlighten us here in this respect?
Regards, Frank Coluccio
|
