Steve, in response to your reasons why DSL will not fair well, I'd like to point out a few things for clarity, and add my views concerning the DSL vs Wireless discussion.
First, Ray J. is right, in that most subscriber pairs do not contain load coils today. Load coils, or "loading pots" as they were once called in a collective sense, were and still are used on longer metallic circuits. Most notably, in the past, between central offices as trunk circuits prior to short-haul carrier systems, and to subscriber locations far away from serving central offices.
Their purpose is not so much (if at all) to suppress higher frequency hum, rather to flatten out the audio response (frequency versus attenuation) of copper pairs once they exceed a certain distance. The inductance in the load coil is used to counteract the natural capacitance of the twisted pair line, and the fidelity of the audio product is improved dramatically (typically from d.c. to 3.8KHz, or 4KHz, or 5KHz or 8KHz, etc., depending on the inductive values of the coils, and their linear spacing intervals).
This is done, however, at the expense of "cutting off" higher frequencies beyond those just mentioned in the parentheses above. But it has little if anything to do with 60 cycle or high-frequency hum suppression, although if what you meant by that was the frequency cutoff I alluded to, then I stand corrected.
But that was yesterday's technology in response to yesterday's needs. Today, these tend to get in the way of ISDN, T1 services, and other higher speed services such as xDSLs and digital loop carrier systems. But they are not the _real_ impediments to deploying DSL today in normal subscriber situations that many in the press have made it out to be.
The real problems, once the distance criteria are satisfied, instead, tend to be dirty copper, i.e., unbalanced pairs, proximity to power line influence and noisy neighboring circuits, pairs with slight leaks to ground, swinging opens, etc.
>>In addition, xDSL will NOT work in any bundle...which terminates in the million plus current DLCs (digital loop carriers).<<
This is absolutely true concerning extant DLCs that were placed into service until recently. The most recent of these loop carrier systems, however, are optically fed by SONET Rings and are beginning to incorporate (or have options to incorporate) DSLAM/concentrator functionality, supported by high-speed SONET feeds from the central office. In these scenarios, they may perhaps be the best positioned of all to deliver DSL, since they will be far closer to the subscriber, as I believe Ray also mentioned, which would have the effect of obviating some of the perils you've so justly pointed out.
Beyond the ability to incorporate DSLAM functionality, I might add, newer (DLC-equivalent or substituting) access technologies are also incorporating DSP techs to support IP and ATM services including voice, complete with SS7 hooks, right at the pedestal point, or at the serving area interface. In fact, WCII's own P-MP employs an over-the-air form of ATM delivery, similar in purpose.
Concerning Winstar's P-MP offering, I think some clarification should be made:
>>As the obvious attractive alternative, the buildout of broadband digital backbones underscores the need for fast, reliable and inexpensive digital broadband local loop solutions. Broadband fixed wireless such as Winstar, provides fiber quality BER, OC3 bit rates (155Mbps, using P-MP technology) to enterprise customers.<<
While I too am eager to see this technology take off, a number of caveats should be mentioned here. One is the need for line of sight, or at least optimal sighting of the receiver with respect to the transmitter. And distance, as in DSL techs, plays a role here too. The other point I'd like to make is that the full OC-3 component is not available at all times to all individual users, just as Cable modems do not avail themselves at all times to dedicated access speeds when others are in contention for the same bandwidth. So, whereas the aggregate speed of the P-MP may be 155 Mbps, the actual allocated b/w to any user at any time may be a function of statistical derivations.
On the other side of the coin, however, WCII's P-MP offering may very well serve as the collection vehicle for DSLAMs in multiple campuses and office parks within the effective range of the transmitter.
I don't believe in all cases it's going to be easy to deploy homogenous technology solutions. I think that's clear to see. I think that tomorrow's networks will be a quilt employing many discrete techs, heterogeneous in nature, as I've begun to point out here, and as FON has suggested, however indirectly, with its proposed ION offering.
Comments are welcome.
Regards, Frank Coluccio
|
