IPTV technology's tough call

Which letter comes after T: N, C or P? For would-be IPTV providers, deciding which acronyms on which to base their networks — fiber-to-the-node, fiber to the curb or fiber to the premises — is only one of the tough questions they face. MOCA or PNA? MPEG-2 or -4? IPTV deployment entails a battery of technology choices that forces each carrier to weigh its short-term and long-term strategies as well as those of competitors and strike the right balance between risk and return on investment.

As this year began, Lehman Brothers analyst Blake Bath bemoaned the high cost of Verizon's FTTP rollout, suggesting that the carrier might consider following AT&T in deploying more FTTN. At the time, Verizon defended its strategy as both a short-term and long-term solution: Short term in that it allowed Verizon to start rolling out video services quickly — using standard cable radio frequency technology rather than the less-tested IPTV — while building in enough bandwidth to keep subscribers happy long after FTTN networks run out of space.

As the year wore on, however, AT&T stirred its own doubt among investors as problems with its Microsoft middleware slowed its IPTV rollout. Verizon has become so frustrated with glitches and delays in Microsoft's middleware, The Wall Street Journal reported this month, that the carrier has been taking on more of the work itself, replacing some Microsoft people and software with its own. Microsoft has argued that the problems to date are par for the course for a major new technology.

Microsoft isn't the only middleware vendor in town, of course. But AT&T reportedly chose that company in large part because it bet that the features Microsoft's middleware enabled would entice cable customers to convert. The trouble with middleware has become a cautionary tale for would-be IPTV providers about the trade-offs between attractive features and reliability. When entering a new industry, some say competitive differentiation should take a back seat for a while.

“On one hand, you want new features; on the other hand, you want reliability,” David Tilley, broadband engineering supervisor for Nevada telco CC Communications, told Telephony earlier this year. “I have never had to reset my middleware server ever. It never locked up because we never asked it to do too much.”

The decision about how far to roll out fiber also involves a complex equation of installation economics and market timing. FTTN may be cheaper to deploy than FTTP (roughly $800 per home served versus $1500 for FTTP, according to Frost & Sullivan), but the longer carriers have to wait for robust, bug-free IPTV service, the more telephony customers they lose to cable guys in the meantime. To date, the major cable operators have about 7 million telephony subscribers, while the Bells have fewer than 200,000 TV subscribers. So calculating return on investment becomes even more complex.

“It's kind of a religious argument,” said Bob Larribeau, director of Multimedia Research Group's IPTV program. “My belief is that deploying fiber is like putting gold in the ground. It's an investment that's going to pay back over years, whereas DSL is an end-of-life strategy for extending copper.

Verizon's FTTP strategy also may apply significant pressure to its cable competitors in the long run. This summer, a CableLabs report warned that, if Internet traffic continues to accelerate, it may end up being cheaper for cable operators to deploy FTTP than to upgrade their networks sufficiently to match the quality of Verizon's FTTP offering. Cable companies have disputed the report's assertion, however.

One short-term way to deal with time-to-market pressure is to partner with a satellite provider for video, as AT&T has done with EchoStar Communications for its HomeZone service — an in-home multimedia networking service that AT&T is deploying where its own U-verse video service can't reach — and both BellSouth and Verizon have done with DirecTV. The satellite services are used to serve multi-dwelling units, where Verizon often uses an architecture similar to FTTC, and as part of a service bundle where fiber-based video isn't available. But as a competitive offering, many regard satellite partnerships as Band-Aids, and some even question how effective they are as in that role.

“I don't think it's an answer,” Larribeau said. Citing the wide gap between the rate of cable telephony subscriber additions and telco satellite video penetration, he said, “The satellite strategy [Bell companies are] using today just isn't winning.”

Copper-based broadband choices also tend to lead carriers to another dilemma in which they must weigh both their long-term and immediate needs: choosing compression encoding technology. Because FTTN and FTTC deliver less bandwidth to the home than FTTP, carriers that rely on copper to the home are motivated to also deploy the newest generation of video encoding and compression technology, MPEG-4, to squeeze their content into as small a stream as possible.Whereas the currently deployed MPEG-2 encoders might require up to 6 Mb/s for standard video and 12 Mb/s to 30 Mb/s for high-definition TV, MPEG-4 can squeeze standard-definition video to about 2.5 Mb/s and high-definition video to perhaps 8 Mb/s. However, MPEG-4 is young and only now making the leap to becoming a mass-market technology. As such, there have been concerns about its readiness and for good reason: Unexpected delays in the availability of MPEG-4 set-tops and their middleware led equipment vendor Tut Systems to miss revenue expectations for the second quarter. Independent telcos also have delayed service upgrades waiting for MPEG-4 set-tops.

While they're in the home, carriers also must decide which technologies to use for in-home wiring. Again, timing plays a key role. Verizon, which had an earlier start in rolling out triple-play services widely, declared its preference for MOCA, the technology promoted by the Multimedia over Coax Alliance, which uses each home's existing coaxial cable for in-home networking. At Verizon's request, vendors have built support for MOCA into fiber access gear. This summer, however, after a year's worth of technological evolution, AT&T chose HomePNA technology, which was originally designed to use homes' existing phone lines but which AT&T now says also can be used over coax, making it more versatile.

A significant factor in each carrier's technology decisions is the level of competition in its own footprint. For example, independent rural telcos may not be losing as much business to cable providers as their cousins in urban areas, giving the rural telcos a little more leeway in their decisions. The Bells, however, are facing fierce competition that is expected to grow even more intense. Cablevision, for example, is deploying DOCSIS 3.0, the latest cable broadband technology, and has begun trials of 100 Mb/s broadband with 15 Mb/s committed rates.

“I called that [AT&T CEO] Ed Whitacre's worst nightmare,” said Multimedia Research Group's Larribeau of the Cablevision trial. “How is VDSL's 20 Mb/s going to compete with a 100 Mb/s pipe from cable? Some people say no one needs 100 Mb/s. In my mind, that's irrelevant because [consumers] understand that 100 is more than 20 much more than they understand how they're going to use that bandwidth.”