Cable Gets Switched On

The cable industry has spent the past several years promising Wall Street that operators’ massive $80 billion rebuild is its last major capital spending effort.

But in the face of bandwidth-hogging services such as HDTV, cable faces a capacity crunch that can’t be remedied through significant new outlays.

With that as a backdrop, the industry has started to look at alternatives to increase channel capacity and use its bandwidth more efficiently.

One centers on the evolution to an all-digital network, which would reclaim valuable analog space. A second choice is to look at more sophisticated rate-shaping and advanced encoding procedures, which will save bandwidth. And in some cases, operators are going ahead with rebuilds, typically with 550-Megahertz systems, hopefully under Wall Street’s radar screen.


But there is another option that’s gained favor in some quarters: switched-broadcast video, in which operators take the least popular channels and instead of broadcasting them to all subscribers, send them out only to those homes where a viewer has tuned in.

Switched-broadcast video gained traction within Time Warner Cable during the era of former chief technology officer Jim Chiddix and through Mystro TV, the MSO’s networked digital video recorder effort.

More recently, it’s been Cablevision Systems Corp. chief operating officer Tom Rutledge leading the charge. He told financial analysts during the MSO’s first-quarter earnings call that switched broadcast video was nearly ready for primetime.

“Switched-video technology will be feasible by the end of this year,” he told analysts in answer to a question about what Cablevision would do if it ran out of channel space.

Sources said Cablevision and Time Warner Cable have been testing switched broadcast video in an effort to develop usage patterns that would determine how many channels would fit into a switched broadcast platform — and thus how much bandwidth could be “saved.”

While the bandwidth savings would be immediate, such savings might not necessarily translate into quick revenue.


But sources said the MSOs that are testing switched broadcast video are weighing the launch of subscription ethnic-programming tiers. Such tiers would be a natural for Cablevision’s New York City-area market.

Those tiers and targeted advertising are two ways switched-broadcast video systems could pay for themselves, vendors have suggested.

BigBand Networks Inc. has staked out the switched-video territory, conducting field trials over the past 18 months. “We think it will be very evolutionary,” said vice president of corporate development Seth Kenvin. “We believe it is commercial-ready. We’ll exit this year with some customers.”


Kenvin said switched broadcast requires several elements. One is the actual switch, which could be placed at the hub or node level.

New software is needed in the set-top box that would request a channel from the switched-broadcast platform. The system also needs some session-control software at the headend to send requested channels to the home.

BigBand’s BMR 1200 has been deployed by operators in the field and could be used for switched-broadcast implementations. “It’s all [Gigabit Ethernet] in and all GigE out,” he said.

The BMR, or Broadband Media Router, is an expandable software platform that typically has been used for broadcast grooming, he said. Switched broadcast would require an expansion of existing deployment equipment, with many more switches.

The key is getting actual viewing data from systems to understand which of the less-viewed channels are the best fits for switched broadcast, said BigBand director of business development Biren Sood.


BigBand is participating in two field trials, he said. In one trial, the MSO is switching 60 channels in a 1,000-home node in a 550-Mhz system.

In that trial, the question is whether switched broadcast is a better financial alternative than a 750-MHz rebuild.

In the other system, the MSO is switching its entire 180-channel lineup in a 1,000-home node.

The goal is to gather usage and demographic data on what homes are watching what channels, said Sood, then determine which channels are the best candidates for switched broadcast.

Here’s how the math might work: A system could determine that it will scale its 180 channels back to 60, putting 120 channels in the switched-broadcast mode.

At 70% cable penetration in a 1,000-home node, that’s 700 subscribers who could do switched broadcast.

If the system has a 40% rate of digital penetration, the true number of homes is knocked down to 280. But since those homes average 1.5 set-tops per household, that’s a total of 420 set-tops for which the system must allocate bandwidth, Sood said.

“You could go with a smaller service group and get a higher level of efficiency,” Sood said. “That’s a financial to bandwidth trade-off. But you have to purchase more switching equipment. It depends on what the operator is trying to achieve.”

For Sood, the key equation is, “What does it cost me per QAM [quadrature amplitude modulator], per household passed for these various alternatives?”

Many MSOs are comparing all-digital systems to switched broadcast, or even considering combining the two.

“It depends on your situation,” Sood said. “The time to market is faster [with switched broadcast] and you can grow it as you need to grow it. You can grow QAM by QAM.”

Sood believes most MSOs could take two-thirds to three-quarters of their channels and place them in a switched-broadcast mode, because 20% of the channels are watched by 80% of viewers.

As to centralized versus edge-switching, Sood said, “We push all the services to the logical edge, to the node level. There is a degree of complexity.

“There are a lot of moving parts. The next progression is to move to the multi-tier switching model and only push to the hub those services being requested. With that, you’re getting some savings on transport.”

But switched broadcast can be much cheaper than an upgrade, Sood said.

“In the 750-MHz environment, they are seeing the crunch coming. Switched broadcast video is a viable bandwidth maximization technique.”

Harmonic Inc. senior director of the convergent systems division Nimrod Ben-Natan said switched-broadcast video is only one of several strategies to free up bandwidth for additional services, including network upgrades, better compression and better rate shaping and encoding.

“People are comparing the three options side by side,” he said. “Each one has pros and cons when it comes to [capital expenditures] and [operating expenditures] and the complexity of operation.”


Ben-Natan is a bit more cautious about switched broadcast, after further study.

“There are some issues with latency, denial service, and how can you guarantee that we have that bandwidth available?” he asked. “How can you minimize latency? What do you do when you tune to the channel and its not available? What do you do with an out of band outage? How do you manage bandwidth to reclaim service?

“And you have to maintain a database and processing control of each subscriber. That’s a level of complexity” that doesn’t exist in today’s cable operation, he said.

Harmonic looks at its digital signaling gateway and GigE transport networks deployed over the past few years for VOD as its switched broadcast-video technology.

“It’s mostly unicast type of streaming,” he said. “It’s very easy and we did it with Mystro. You can carry on-demand and switched-broadcast service in the same transport.”

Ben-Natan said Harmonic’s DGS will support switched-broadcast video, but “there is a need for client software in set-top and software that maintains the connections to the content source,” such as the signaling software used for VOD-session setup.

“The bandwidth resource management system needs to be open,” he said, so outside vendors that do policy-server work, such as Camient Systems, could play in this space.

Ben-Natan also offered another strategy. “Take the feed from the encoder and instead of multicasting it, load it onto the server and make it an on demand channel.

“You can reuse the same architecture. Video servers will actually record all those channels. You may need more edge QAMs,” he said, but the setup could be economical, especially as storage costs continue to drop.

Terayon Corp. has its CherryPicker technology in the field, which could be used for switched broadcast video. But the company believes switched broadcast video is only one part of the bandwidth-saving business model.

Operators also could look at targeted advertising.

“The same infrastructure will be needed for targeted advertising,” said Terayon Communication Systems Inc. vice president of video-architecture technology Michael Adams. “It’s a similar infrastructure for VOD and telescoping ads.

“By revising the switched-digital broadcast architecture, it is possible to use a common resource-manager model to also support targeted and telescoping ads, VOD and all-digital,” Adams said. “The advantages are that more QAM channels are included in the pool of resources and the statistics improve.”

But there are a number of issues that confront operators in looking at switched-broadcast video. For instance, Adams said there’s a question of how much channel blocking or latency is acceptable.

The first time a consumer turns to a channel and can’t receive it could yield an angry phone call.

There is also the unanswered question of surfing through a lineup of switched-broadcast channels. Tearing down and setting up sessions could be even more complex than anything related to video-on-demand.

Operators may want to think through the evolution of the existing QAM and Internet-protocol transport schemes, and how that relates to switched broadcast, Adams added. VOD, switched-broadcast and targeted ads could go all the way through on an IP-transport network, he said.