Mike,
From my viewpoint, we're not looking at a monolithic set of changes or migrations in
the cable industry, as would be characterized by lock step migrations. Rather, the
industry is still very fragmented and we're witnessing many different directions of
change within the same industry. HFC, however, seems to have finally taken hold and
while each operator or MSO may elect to configure its own internal core network and
distribution lines differently, they will still be bound by a universally relevant set of
attributes and interfaces, since they must all hand off traffic from one to the other, and
to the greater global network. And the latter, itself, is undergoing changes at the same
time.
>>But don't forget, any networking ATM switching equipment could have been used
instead of Fore's. I don't believe Fore's ATM switching equipment was purpose-built
for HFC. Also Tellabs does have voice over cable expertise. And remember Ascend
bought Stratus (SS7 telecom support) lately. How this integrates into an HFC solution,
I'm not sure? But I think it does. Maybe DenverTechie or Frank knows? <<
You've made several points. In some ways, cable operators' head ends will resemble ISP POP server farms as newer services get unpacked and ready for delivery. In those situations where the operator will continue to force feed switched services over DSOs over spectrum using classical FDM and OFDM, they will resemble the telco end office switch locations we have all become familiar with. But in a very short time frame, I think that these operators will live to regret that selection of investment, since the fast emerging IP alternatives will permit a faster ramp-up to application level convergence at the protocol [namely IP] level.
ATM is not an inhibitor in this sense, since it is poised to support IP as well. But it does add another level of infrastructure and complexity which is increasingly falling out of favor with architects, albeit of the bleeding edge variety, as optical solutions begin to unfold, and as higher capacities of bandwidth become available through DWDM and optical switching & routing techniques.
I suppose the benefits and the costs associated with ATM in FTTC architectures need to be examined on a case by case basis, still, since time to market is a factor and some of the emerging techs which are still just over the horizon are not ready for prime time, and may not be for another two or three years in a way that will scale well in existing subscriber networks. Backwards compatibility is a bummer in this regard, requiring dual operations platforms to be maintained, and most operators are not up to this without massive capital infusions to handle the costs. This presents grounds for argument in favor of the clout that T affords in the TCOMA situation.
Getting back to ATM, if it is embedded in the core of a network, and extended out to the neighborhoods, then it's almost a no brainer that it will be used to the premises as well [in the case of the FSAN and other FTTC architectures such as BroadBand Technologies and Next Level]. Despite the conventional wisdom associated with this tact, there will always be newer alternatives becoming available, and that means that the ATM solution may well be relegated, in record time, to legacy status against which new forms of optimization may be motivated. Very Fast IP flows come to mind, for not only voice and data, but for video as well over time. This will require price points for routing to come down further, but isn't that a common factor that has to be taken into account? I am encouraged to read that some emergent companies are beginning to focus on optical level routing. One such company is talking about "bar coding" optically based streams over discrete wavelengths for this purpose. It's going to happen sooner than we'd commonly think. Unless, of course, the rush into now-current techs forces another five to ten year depreciation lock on progress. It happens.
I'm not specifically aware of any direct involvement between SRA and cable telephony, although those technologies are being ported by others as we speak, and I see no barrier to ASND's ability to do the same. But a quick glance at the ASND web site reveals that they have been primarily, up until now, devoted to the traditional carrier model in this regard. From their web page that describes their multi voice strategy ascend.com , I offer:
----
Introduction
Ascend MultiVoice
Transparent Connectivity
Access Port Diversity
Exploiting the Power of Ascend Switches
Three-phase MultiVoice Roll Out
Phase One-MultiVoice for IP and Frame Relay
Phase Two-MultiVoice over ATM
Phase Three-MultiVoice Platform
----
That's not to suggest that they don't have something on the burner ready, for release tomorrow, however.
Regards, Frank C.
|
