It would be neither unfair nor untrue to say that the explosion of high-speed content and services coupled with bandwidth-consumptive high definition caught nearly everyone in the telecommunications space by surprise. Networks built just a decade ago with enough room for a speeding Hummer have become more dangerously congested than the Los Angeles freeway at rush hour.
Cable is searching for ways to salvage cleverly and flexibly built hybrid fiber/coax (HFC) networks that appeared long-term safe on paper 10 years ago but looked a little shaky five years ago and today are definitely wobbling.
"Consumer consumption and the contention for narrowcast services has increased and we find ourselves having to do some node splitting," said Ken Williams, director of engineering at Cox Communications.
Nodes--the places where the fiber ends and the coax begins--were built to be split as demand increased. Some large 1,000-home nodes, in fact, were built to deliver quick efficiencies with an eye to a future when they would be cut down.
Upstream is "the easy split," Williams said. Downstream, where the biggest demand for traffic is greatest, poses a bigger challenge from both a technological and, most importantly, financial perspective.
The best way to split these nodes would be to simply run a second 1310 nm fiber to carry extra traffic then run each fiber into a split node. Of course when you get to that point, you might as well just rebuild the whole network so it's not really a solution for cable operators who are watching their cap ex in a tough economy.
Still, the nodes have to be split to more efficiently deliver traffic so Cox issued an RFP to determine the best way to do it.
"There was rally no easy way for them to subdivide and go fiber deep with the fiber network that was in place," agreed Mani Ramachandran, InnoTrans' CEO. "Our innovation was the development of some new technologies that allowed us to do multiwave transmitter technology at a price point that is suitable for cable TV."
This innovation proved especially effective for Cox, which has built redundant ring-in-ring networks. Cox uses one laser for both rings and then splits the optical signals so that the two paths arrive simultaneously at the same destination, Williams said.
This is, he said, an effective way to keep the network up and running and reliable. The nature of the rings-one is generally longer than the other-gums things up when trying to further subdivide the traffic with the most commonly used method; pre-distortion, Williams said.
"You have to know a lot of things about your fiber: the distance, the type of fiber and so forth so you can pre-distort so that it arrives at the destination compensated," said Williams. With an asymmetric ring-in-ring architecture, "Which side do you pre-distort?" he asked. "That solution didn't work for us."
Cox, Williams said, "asked all the optical manufacturers to help solve this problem and we got a variety of responses, but the only one that fit for us was InnoTrans."
InnoTrans uses 1550 nm technology to carry multiple wavelengths across the same fiber so Cox can separate its content and push it onto separate wavelengths, not actual fibers. The separated traffic can then be carried over the same fiber and dropped off at different coaxial termination points within the same nodes. This effectively splits the node sizes by creating different delivery points but does not physically alter the HFC configuration.
"We allocate the different signals to different people by saying 'This wavelength goes to this set of customers and that other wavelength goes to another set of customers,'" Ramachadran said. "With our technology the distance does not matter. We don't have to optimize for a specific distance."
InnoTrans technology multiplexes the signals onto the existing fiber and makes sure it can un-multiplex (if that's a term) at the end point. While technically still part of the node, that end point becomes, in effect, a smaller node, the InnoTrans boss said.
"Sometimes we think of nodes as physical entities and being physically separated; sometimes they are the same housing, very similar housing, very close together," Ramachandran noted. "The fiber assets that cable TV has in its networks is a very valuable technology and not something that is very well utilized today by just putting one 1310 wavelength on the downstream and one on the upstream."
Williams agrees. He also agrees that building out more fiber isn't a viable option for cable operators who convinced both themselves and the moneylenders that their networks were set in stone.
"We're trying to use the same fiber assets," he said. "InnoTrans really solved that problem for the industry (and) that's our standard going forward until another vendor comes on the scene with an acceptable solution as a second provider."
Unlike Cox, which for now is restricted to a single vendor, InnoTrans is not bound to a single operator. It's providing its solution for both residential and commercial plays across the MSO space.
"We have other top MSOs who are using this," Ramachandran said. We also have deployments and some trials going on with Comcast' we have deployments in Shaw Cable and other Tier 2" cable operators.
The work, he said, is built on questions that Cox asked and InnoTrans answered.
"Cox is more forward-looking in terms of their planning and how they want to do this," he said.