I'm still trying to assess if there's any meat on OpenNET's bone, or if it itself was just a bone thrown on the table to stall T/TCI a couple of months ago. Their most recent news release that speaks about gaining momentum ("OpenNET Coalition Gaining Momentum And Members") is almost two months old.
A seldom spoken truth is that the current HFC plant (including the head ends) are not conducive to a drop ship kind of implementation of secondary (and beyond) ISPs on the same wire. It's not the same as an ISP throwing up an access mux in the central office in the telco realm. Its far more complex to give multiple service providers control of their own destiny on cable plant, if each is to have some control over future DOCSIS features and functionality at the subscriber's residence, including network management and remote provisionsing. I've wondered why no one ever mentions this in the press.
[I guess we'll see some mention of it now. I've inferred that quite a few journos and gurus du jour read this thread for ideas. It's got to do with orwellian feedback loops, the kind recently discussed in the last mile thread.]
The complexity increases geometrically with the number of additional providers when you introduce voice services (what I like to call cable-iphony, playing off the doulbe entendre qualities of IP and "if," using some license, of course). And if anyone thinks that ATHM is not being pressured into a voice role, then they have not been paying attention.
As recently as last fall Malone and other MSO heads were contemplating the Packet Cable implications (ramifications?) for those cable entities who did and did not participate. Malone actually stated that Voice was inevitable, and it could be achieved over cable plant over a switched paradigm, or one that was routed. He went on to state that if he had his way it would be routed (over IP). The only ISP-like entity at his immediate disposal capable of administering this would be, guess who? @Home.
And if T/TCOMA chose an alternative approach, then they would be forced into creating a dichotomous situation. In such a process they would lose the argument which I present below. The argument states the difficulty of doing this for others, thus rationalizing why there could only be one voice or ISP per physical cable distribution system.
The Packet Cable model which supports voice over IP would establish a separate telephony environment (similar in makeup to the PSTN) for all participating Packet Cable service providers. These providers would reside on a cloud which would support 'on net' connections from one cable subscriber to another, irrespective of which cableco they subscribed to, as long as the provider was Packet Cable compliant. Anyone wanting to dial "off net," e.g., onto the PSTN, could do so through a Packet Cable-to-PSTN gateway which would be transparent to the user.
What I've just described is currently being worked out by Cable Labs and the participating MSOs. If multiple independent ISPs were permitted onto an MSO's plant facilities, directly, the complexity of this would rise proportionately, or to the point where the condition was untenable over a single bidirectional cable plant split.
The apparent solution (to me, at least) in this instance would be multiple physical cable paths, or to structurally separate the physical medium of a single path (as we now have) from the upper networked layers. In the latter instance, the MSOs would be relegated to mediation of the distribution of Layer 2 and Layer 3/4 presence on each system, just as the telcos now administer ISP access via dialups, only in a way that is far more complex.
This would amount to rendering the cable folks as physical layer utilities only, renting address space (spectrum/QoS?) on their systems to all comers. The MSOs, and now T, are not about to entertain this at this time, for obvious reasons. It's got to do with a golden goose. Comments welcome. Frank_C. |
