Hello Doc, yes there are many topics buried in the implications of your
uplink post. We know (or think we know) what we have today. Let's look
a couple of years out. Anything shorter than two years out will be purely
tactical, as in warfare, and subject to the issues I cite below, anyway.
From a 60,000 feet perch I see T encroaching on all of the ILECs
territories with potentially massive build-outs of Lightwire and wireless
over time, as an improvement on their HFCs and still-to-be-converted
all-coax plant. This, in itself, even if it were only limited to cable TV and
Internet services, would be threatening enough. But T will hardly stop
there.
T is going for the whole schmear, the entire enchilada, by deploying
voice and undoubtedly so-ho and work-at-home data services in the
future, as well. Note that none of the latter enhancements are currently
supported in a broad manner by the MSOs, much less ATHM, at this
time. And in so doing, they become eligible to pick up millions of VPN
subscribers who straddle their work days between home, office, and the
road, which heretofore have been the province of dialup and ISDN
providers, only.
Of particularly high concern should be the thousands of enterprise level
VPN contracts which are signed annually, some of which range into the
50,000 to 100,000 partial work-at-home user range. Today, cable modem
users are, for the most part, precluded from this class.
T truly means to provide a complete "bundle" which is near-
all-encompassing, save perhaps for some uniquely satellite-supported
services. But they still have their options open in this respect, too.
The ILECs have thus far, in turn, responded by using a technology that
was initially intended by the BOCs in the early part of the last decade
(wow, that really makes it sound old, doesn't it?) for switched digital
video (SDV) and video on demand (VOD) services. And here, of
course, I'm referring to ADSL.
As a work-around to some of the limitations of ADSL -- such as its
distance and speed limitations, and availability in remote serving areas --
some of the ILECs have chosen to offer substantially "purer" forms of
voice and data (particularly motivated by Internet) delivery through the
use of integrated fiber in the loop (IFITL) architectures.
ITIFL borrows from the original passive optical networking (PON)
schemes which were established in the full service area concepts
(FSANs) by various ITU working groups and Bellcore about a
half-decade ago. The following posts here in SI address BLS's version
of IFITL.
techstocks.com.
techstocks.com.
I suspect that the orange "inner-ducts" you saw, which are usually
fitted three to a larger 4" conduit, are being placed in preparations for
introducing the latter. I.e., a variation of the IFITL overlays I cited
above, or some other means of extending fiber to the neighborhoods and
attaching to a Next Generation Digital Loop Carrier scheme in
compliance to Bellcore specifications. This is what all the speculation
was concerning AFCI a short while back. In the case of BLS they used
a Marconi solution. These may be different at the board and component
levels, but they essentially will be able to deliver the same types of
services.
The need for additional field-level "buildings" and central offices, or in
the case you cite, FCOs, diminishes with fiber due to its distance
transparancy properties. Services can now be "backhualed" over
enormous distances because of the low loss and high bandwidths
permitted. [I recently wrote about this in the FCTF while speaking about
how storage attached networking is now possible over far greater
distances, as opposed to being limited to "inbuilding" deployments as
they were in the past, through the use of dark fiber, or fiber which has
been lit and provided as a "managed service." But I digress.]
The main point here is that there is no need for creating fiber central
offices purely for reasons having to do with proximity, or because of
other architectural reasons which were binding in the past.
There may be other motivations, however, about which I could only
offer some speculation. One possible reason would be to establish
localized caching and leveraging that location for plant purposes, but
even here, fiber's vast bandwidth potential could easily be used to haul
content over great distances. Perhaps a more compelling reason in the
future which would be used to establish FCOs, or some other functional
type of building deeper into user territories, would be to support actual
web (or even individual user) hosting services.
Even where traditional MSOs are concerned, today, we see one master
head end satisfying the distribution needs of many neighborhoods,
where each of which required their own head ends in earlier times
because of franchise boundaries and for reasons which stemmed from
shorter black-cable (coaxial) reaches, price-performancewise.
Variations of these principles exist where a "master head-end" might
now be used for distributing to a number of smaller, clustered
distribution-level h-e's, in a top-down tree distribution manner. A lot of
the present topographies have resulted from consolidations, clustering
activities by the MSOs effected by "swaps," etc. And some topologies
are circular and ring based, deriving from SONET's counter-rotating ring
attributes. Newer systems will leverage fiber's transparency, however,
further limiting the number of actual "structures" which will be required
per unit of homes passed.
=====
A salient point which emerges here is that the pure bulk and girth of service provider physical plant is declining radically, while the amount of bandwidth they will be delivering is set to soar by orders of magnitude.
- Massive amounts of twisted pair cabling and its associated field infrastructure is being replaced by several strands of fiber in the feeder and distribution plant.
- Central offices in some regions are closing their dorrs, shutting down, as a result of consolidations which are enabled by the distance transparency effect afforded by fiber.
- Form factors of network elements and entire platforms are getting smaller with increases in chip densities and integration.
- Object switching in software based architectures is replacing many of the network elements and platforms that were once supplied only in hardware.
What this all spells out to me is what Jay Lowe and others have spoken about here and in the FCTF in the past, and that is that the network itself will become inconsequential at some point, and all that will matter is the distribution of coded objects over enough bandwidth. At some point, Negroponte's analogy to "atoms and bits" begins to make a lot of sense here.
------
Once the MSOs and the ILECs begin to reach these levels of bandwidth abundance and transparency, it will then be possible to "bridge" between them where it makes business sense to each of them, using gateways with a minimum of fuss. This, where the digital content is concerned, in any event. The analog video component of the MSOs' architectures is another story, which I address again, in a moment.
Bridging is already possible to some degree (for the purpose of swapping and peering), but the MSOs have far more bandwidth already deployed for non-voice services than the ILECs, so this would make any form of equitable trade scenario between their network's edges and the subscriber seem unlikely right now.
Also, the ILECs have not gotten their traditional video agendas straight yet, after their earlier setbacks in this space, nor is it certain that they ever will. They may jump, instead, directly to digital. With sufficient bandwidth deployed, they may again seek to circumvent the CableLabs genre of cable video and come up with a switched digital video approach of their own based on PON delivery, perhaps this time one that will actually work.
The greatest limitation to a swap scenario lies in Cables continued use of RF design in the last thousand feet. Here they present an RF format that I doubt the ILECs will ever return to in force, thus introducing a dissimilar line format which represents an incompatibility to free exchange (while the ILECs last thousand feet will in all likelihood be all digital).
======
What does this all portend for ATHM? The above is based on several years of buildouts. Where will ATHM be situated then, what will its exclusivities be, and what will they want to do for a living then? And how beholding to ATHM will T and the other MSOs be by that time? I don't know the answers to any of these questions. Do you?
I'm tapped. Back to you and others for comments and corrections.
Regards, Frank Coluccio |
