Bridging the MPEG Gap
Company claims first IP-MPEG video converter
By Karen Brown
from the May 21, 2001 issue of Broadband Week
Internet and traditional broadcast video formats are two distinct worlds that up to now have been separated by a deep technical gulf. Without a bridge to connect the two, cable operators can't easily funnel bandwidth-efficient Internet video content directly into interactive digital set-top boxes.
Not surprisingly, system providers have been racing to build that bridge, and now Yorba Linda, Calif.-based Coax Corp. is claiming it has crossed the finish line first. Coax says it has developed the first IP-MPEG bridge system, able to knit an Internet backbone directly into a hybrid fiber-coax cable TV network.
The system not only will enable IP video content from the Internet to flow into existing digital set-top boxes, but also will open the way for any mix of Web and broadcast applications an operator wants, according to Jeffrey Fry, chief business officer at Coax.
"We have made a universal platform for a digital headend," Fry claims. "It's like a universal media translator. You can put any type of data or video or anything on one end and get it out in the form on the other end to transmit on either cable, fiber or twisted pair."
The Coax product consists of a digital controller box for the headend and related software components able to simultaneously receive IP and MPEG video and then transmit it in standard broadcast MPEG-2 video to existing set-top boxes. The bridge actually is a scan converter that takes the Internet-jumbled IP MPEG data packets and re-assembles them for MPEG-2 delivery down cable pipes. A generous buffer allows the unit to adjust for dropped packets, so the end result is a standard broadcast video.
That, in turn, allows video served up by Internet video servers from the likes of Microsoft Corp. and RealNetworks to be viewed directly on TV, a PC or a wireless device.
"Everybody has become parochial," he says. "With real convergence, it doesn't matter what you are sending down your pipe. And that's what we're driving for. We're making a universal translator, so who cares?"
As such, Coax says its system will be flexible enough to add new video formats--including MPEG-4 and the future MPEG-7--with a software upgrade to the cable headend controller. Fry says Coax will make the controller middleware specification open, so developers can add a software module to support these new formats.
That flexibility does require ample processing firepower, and Coax's controller box gives each QAM modulation QPKS card controlling the upstream and downstream data a 450 MHz processor of its own. Fry claims this will not significantly drive up the cost of the controller because processor prices are dropping.
Asking a cable operator to invest in a slew of new controllers won't be easy. But Fry argues that Coax's universal video controller and IP-MPEG bridge will make coax-based video transmission more efficient, thus easing the need to push expensive fiber closer to the subscriber.
"The only way they are going to invest is I've got to show them a lower cost migration path to convergence," he says. "What I'm trying to show them is, 'Hey, I'm showing you an intermediate step between coax and making everything fiber that allows you to get revenue in the interim.'"
While talks are underway with the major MSOs, Coax is finding early hooks with the smaller cable operators. Its first deal is with Texas cabler Buford Media Group.
But he also acknowledges future success relies on forging deals with the larger MSOs. Coax is in involved in informal testing and talks with AT&T Broadband, Comcast, Time Warner Cable and Charter, Fry says.
Such an IP-MPEG bridge would be useful, but the real question is whether the cable operators need it desperately right now, according to Branko Gerovac, vice president of research at SeaChange International. SeaChange is jointly working with Microsoft Corp. to meld MPEG broadcast functions onto an IP network, with the goal being a fully IP system.
"It's a question of how the migration path occurs," Gerovac says. While he believes IP transport will become important for entertainment and data delivery on all types of data networks, cable operators "are not going to need to move to IP very quickly, because they haven't stressed the capacity of the HFC network currently. So they have a lot of headroom in delivering advanced services over the next few years. But eventually it will be very compelling to them to move to an IP transport environment, and the question becomes how you end up doing that."
If cable operators suddenly found themselves cash rich with big demand for IP-based services, they could order a full plant upgrade to IP with no transition. It's a safe bet that won't happen, so Gerovac thinks it more likely there will be a gradual transition, with operators inching IP slowly into their systems.
"It depends on the individual cable operator, as to what kind of scenario is most appropriate for them," Gerovac says. "We'll actually see something between those two scenarios. I think that we will see a package of services that require an incremental upgrade to IP. And I think that also, IP has a history of sort of getting in the back door in an communications environment, and I think that we will end up seeing that, where all of a sudden you will start seeing IP getting in the back door and it will just appear."
|
