Last month, Time Warner Cable said it planned to roll out switched broadcast video in several markets this year, with full MSO deployments scheduled for 2006 and 2007. The company said the plan grew out of successful tests in Austin, Texas.
More recently, during the Society of Cable Telecommunications Engineers Cable-Tec Expo in San Antonio, Time Warner executives provided some details about what went on in the test. They described the problems they faced and offered an outlook for bandwidth savings from switched broadcast.
Paul Brooks, senior network architect in Time Warner’s advanced technology group, outlined the Austin test in a technical paper delivered at SCTE’s annual conference.
The test started out in a single node with 334 homes, then expanded to 1,550 total set-tops.
A small software client was loaded into the Scientific-Atlanta Inc. set-top boxes, Brooks said. Participants in the trial had no idea some of their channels would be “switched.”
That was intentional, Brooks said, because the goal of switched broadcast is to make channel changing invisible to the consumer.
He said that in future rollouts, the switched software client would be embedded in S-A’s SARA navigational software.
When consumers changed channels in the test, the newly downloaded set-top software sent a viewing request to the hub.
A BigBand Networks Inc. switch in the hub processed the query and switched the requested channel into the home.
Time Warner hub locations serve about 20,000 homes, Brooks said. (Time Warner will use a generic Ethernet switch in future launches. A request for proposals for that switch has been released.)
The software client inside the set-top tracked and reported user activity, including volume and mute control.
One issue with switched broadcast video is determining when to turn the stream off, Brooks said. A subscriber who leaves a room and leaves the TV on all night, for example, “wastes” bandwidth.
By monitoring “volume” and “mute,” Time Warner can tell if a stream is still active. The company decided to send out a message after four hours of no-volume activity to check and see if a viewer was still there. A prompt appears on the screen: “To continue viewing, press Select.”
Some 170 channels were part of the test. Only digital channels were switched. Analog channels, VOD and HDTV offerings were not part of the test, Brooks said.
Programs were encrypted using S-A’s PowerKey software. The networks that were being switched were put in the same quadrature amplitude modulation units, to maximize bandwidth efficiency, Brooks said.
Brooks said Time Warner had a number of concerns going into the trial — but found that potential problems didn’t necessarily materialize.
“Channel surfing is not an issue because the switch was so fast,” he said. Subscribers couldn’t tell their channels were being switched, he said.
“Subscriber privacy was maintained,” he said, because set-top addresses were scrambled prior to logging on and log records couldn’t be traced to consumer accounts.
The company did find some non-responding set-tops and rolled trucks to fix the problem. But those issues were not associated with the switched technology, Brooks said.
For SBV to work, Brooks emphasized, the plant must be clean — especially the upstream portion that consumers use to change channels.
The system did find a few upstream problems, he said, but they were confined to home-wiring issues.
The number of peak streams viewed was 110 of the 170 channels for the 1,550 homes. The larger the sample size, the larger the total number of channels viewed: At 500 homes, there were fewer than 80 channels viewed.
In future rollouts, Brooks said, the channels to be switched will be limited to digital simulcast channels and live pay-per-view sports and events.
It will be tough to handle both switched broadcast video and video-on-demand because their peak usage tends to occur at the same time, Brooks said.
Evenings and weekends see the most switched-video activity, which isn’t surprising, because that activity tracks regular TV-usage trends. VOD follows the same pattern, he said.
“Switched digital broadcast is real and works on existing set-tops,” Brooks concluded. “It can be cost-effective when properly deployed. Operational channels can be met and valuable viewership data is a side benefit.”
Operators can save 12 to 15 channels of 6 MHz, but it’s not for those who want to take baby steps. “Go big or go home,” was his concluding advice.
One problem Time Warner didn’t run into in Austin, but will face in other markets, is that of customers with one-way CableCARD TV sets.
The MSO has about 6,000 such subscribers who wouldn’t be able access switched-broadcast networks because there is no return path from their sets. Time Warner engineers say they’re working on options to solve that problem.