Cable Industry Redoubles Efforts To Define Next-Gen Architecture

Mike LaJoie, CTO, Time Warner Cable

Mike LaJoie, CTO, Time Warner Cable

December 14, 2009 – At a level of intensity and consensus not seen of late the cable industry is rapidly moving to coordinated development of an HFC network evolution strategy that taps IP technology to deliver any type of service to any device without restrictions on time or place.

Based on input from a variety of industry leaders, including the new president and CEO of CableLabs, Paul Liao, several aspects to the new strategy are discernable, starting with the fact that the goal is to vastly expand capacity to deliver on-demand content without requiring extension of fiber all the way to the premises. “That coaxial cable that runs outside your house can carry five gigabits per second easily to your home if we harness it correctly,” Liao says. “And that’s an evolution. It’s not laying new fiber or new plant.”

The shift to on-demand programming beyond the traditional VOD approach, now underway with trials of TV Everywhere serving content to subscribers’ PCs and evolving eventually to network-based DVR services, has become a given of operators’ strategic thinking. “We all like the network DVR idea,” says Comcast Cable CTO Tony Werner.

Werner notes that VOD and random access to content “are at the tipping point” in light of studies showing that “close to two thirds of viewing hours are spent accessing content via DVRs or on demand.” And that doesn’t factor in the Web, which Werner characterizes as the source of “the biggest impact” on entertainment consumption.

“With the explosion of throughput, storage, super-fast connectivity and devices the world is going to look very different five years from now,” says Mike LaJoie, CTO at Time Warner Cable. This means that, along with the rapid move to dedicated unicast delivery of on-demand content, where the capacity-conserving advantages of shared access to broadcast channel streams don’t apply, cable operators must also adjust to the tetherless nature of consumer behavior.

“The first thing a person reaches for in the morning is a personal device that isn’t plugged into the wall,” LaJoie notes. “Cable owns a great big plug. We have to figure out how to work around that.”

These two big elements central to evolving HFC architecture – unicast capacity requirements and achieving an efficient way to get that unicast content to any device in or outside the home – comprise the heart of the new network migration strategy taking shape at CableLabs. Its hallmarks include abandonment of the notion that fiber-to-the-home is inevitable and an embrace of IP as the eventual platform for delivering all services.

Where the capacity issue is concerned, after years of debate the industry appears to have settled on an enduring commitment to coaxial cable as the last-mile conduit in most circumstances. Two years ago CableLabs caused a stir in industry circles when it issued a white paper suggesting fiber-to-the-home would be essential to meeting future capacity requirements. But today Liao disputes the notion that there won’t be enough bandwidth to meet industry needs and competitive challenges without an all-fiber solution.

“The question to ask is, who really needs that bandwidth?” Liao says. “Already consumers can get 8 to 10 to 15 megabits per second, and with DOCSIS 3.0 we’re talking 50 to 100 mbps. To my knowledge there’s no application that needs that. Even HD can be done with a few megabits per second.”

“Of course,” he quickly adds, “we can drive fiber deeper into neighborhoods if that’s necessary.”

Sources report that, in conjunction with that deeper push of fiber and elimination of in-line amplifiers on the coaxial plant, cable strategists are taking a serious look at the notion of a last-mile infrastructure that resembles PON (passive optical network) but without the fiber. This could be accomplished through use of an environmentally hardened external distribution gateway which would convert amplitude-modulated optical signals on the fiber feeder to RF for passive distribution of the total payload to all households. Devices would access dedicated content based on the addressing mechanisms of the IP distribution protocol.

Achieving capacity efficiencies will also require use of H.264 MPEG-4 compression, which adds another incentive to relying on IP as the convergence platform. MPEG-2 has been a big improvement over analog, “but it’s still wasteful compared to the way we could do it if we went to a really all-digital IP world,” Liao says.

High-end video services such as super-HDTV family theater VOD and 3-D programming might provide a pay-as-you-go approach to installing next-gen set-tops, all of which operate in hybrid MPEG-2/MPEG-4 mode, thereby easing the transition to all-IP. For example, current CableLabs thinking is that a Blu-ray quality 3D HD service will require an additional 50 percent bandwidth increment over the bandwidth allocated today for HDTV. But, if MPEG-4 is used, that would actually result in a reduction in the bit rate compared to MPEG-2 bit rates for HDTV. “I think there’s a path forward on that,” Liao says.

While MPEG-4 is a major driver behind any move to IP convergence, a transitional step involving use of next-generation hybrid set-tops may turn out to be a short-lived component of the overall migration strategy if new thinking at Comcast prevails. Sources say the company is looking at its all-digital strategy employing low-cost DTAs (digital terminal adapters) in analog households as a model for moving to all-IP. In this architecture the heavy use of IP-based network intelligence would eliminate the need for super-smart home terminals.

In truth, the capacity issues and possible solutions are a subset to the real driver behind network evolution, which is the need to support converged services. “The Internet has really changed everything,” Liao says. “And we will see the convergence of cable systems toward utilizing those technologies to really revolutionize the cable experience.”

Liao says he has “challenged the folks at CableLabs to really move in that direction. I think this convergence of anytime, anywhere, any place and on any device is really the focus of what we’ll have over the next couple of years. And all of that will be driven by the fact that the Internet, the IP protocol, makes all that possible.”

A good place to start is with agreement on new approaches to DOCSIS 3.0, content protection and content authentication in the context of getting TV Everywhere to scale industry wide. “All of those things are in play, and all of those things are important,” Liao says.

“One of the deficiencies of the industry is that there are so many different cable MSOs following different approaches, and unless they come together on a common specification, common standard, they can’t have that scale,” he adds. “In today’s competitive world scale is very, very important.”

Another area of uncommon ground has been individual MSO approaches to VOD architecture. “Mike [LaJoie] and I have been talking about how we can get more coordinated on a couple of platforms,” Werner says. “Making on demand more common will help.”

“Time shift is going to happen, and the trick is how to maintain and promote it and make it valuable for the consumer,” LaJoie says. “TV is the killer app. What else would people sit and stare at for four hours straight?”