Warning: Highly technical post ahead. I know it won’t appeal to some readers, but I got answers to questions I’ve been curious about for a while and thought they were worth sharing.
If you follow the cable space, you may be aware of the debate around how MSOs can best deliver IP video. The push has been toward a modular CMTS (cable modem termination system) approach. However, Motorola recently proposed a different method called the DOCSIS IPTV Bypass Architecture, or DIBA.
DIBA has been covered by several trade reporters, but most recently Brian Santo at CED wrote an in-depth article on the pros and cons. His story brought out some information that was new to me, and so I thought I’d get some clarification from Motorola’s Jeff Walker. Below are my questions and Jeff’s answers. (Thanks, Jeff)
Q&A with Jeff Walker
Q: Explain the difference between routing IP video through a modular CMTS and routing it directly to an edge QAM.
A: In a modular CMTS approach, IP video travels to the M-CMTS core, then heads back through the Converged Interconnect Network (CIN) to the Edge QAM. With DIBA, however, IP video travels directly to the Edge QAM, requiring fewer M-CMTS core components and less CIN switching bandwidth. DIBA’s direct approach scales much more efficiently as IPTV traffic grows.
Q: Motorola makes CMTS. Why are you proposing a solution that might cut into your equipment sales?
A: Because we believe it will provide operators with an efficient and cost effective way to support IP based video delivery. The growth for CMTS equipment will continue to expand as MSOs offer higher bandwidth services based on DOCSIS 3.0 channel bonding. (Note: the reference here is to higher-bandwidth data services, although that includes things like video delivered over the Web)
Q: Why are you proposing that DIBA be added to the DOCSIS 3.0 specification given how far along the spec already is? (Note: This question clearly hit a nerve.)
A. We are not suggesting this at all…this not true…we suggested that it be taken up AFTER DOCSIS 3.0!!!!
Q: Cisco and BigBand Networks have criticized the DIBA approach saying it adds complexity and doesn’t save money when you look at the total cost of ownership. How do you counter those arguments?
A: It actually reduces complexity and saves significant amounts of money because operators don’t need to buy additional CMTS equipment and Ethernet switching (as explained in answer #1) to support IP based video delivery.
Filed under: Cable, DOCSIS 3.0, IPTV
[...] instead the chart below (click to enlarge) on how much bandwidth specific applications require. Motorola’s Jeff Walker presented the chart in a discussion of TelcoTV at, ironically, the Cable IPTV show. This chart [...]