DOCSIS 3.1: What's Next?
DENVER -- Cable Next-Gen Strategies & Technologies -- It's been five months since CableLabs released the DOCSIS 3.1 specification, and a panel of speakers here, moderated by our own Alan Breznick, had a lot to say not only about the progress of development, but also about how the industry needs to think about the next DOCSIS spec in the context of the many technologies that will enable the next generation of cable infrastructure.
DOCSIS 3.1 is competing with many other technology priorities, including the Cable Converged Access Platform (CCAP), Remote PHY, and explorations of software-defined networking (SDN), just to name a few. However, Dan Rice, senior vice president of network technology at CableLabs, said the entire purpose of the new DOCSIS spec is to help operators "stay ahead of organic demand" in a very practical way. DOCSIS 3.1 will get deployed only when, where, and how it makes economic sense, and it will be only one of many tools in the network toolkit. (See What’s Next for Cable Tech?)
From a practicality standpoint, DOCSIS 3.1 has the potential to deliver capacity of more than 10 Gbit/s in the downstream and 1 Gbit/s in the upstream, but that's not what will happen first. Jorge Salinger, vice president of access architecture at Comcast Corp. (Nasdaq: CMCSA, CMCSK), said cable operators will have to get rid of traditional video before the necessary spectrum is available for that type of magnitude shift. "Eventually, we'll have IP video as the only transport… but it will take a long while."
In the meantime, cable companies will be able to take advantage of DOCSIS 3.1 without making any initial changes to the cable plant. Then they'll be able to reclaim bandwidth gradually for further utilization as needed through strategies like the migration from analog to digital cable.
As far as the timeline is concerned, DOCSIS 3.1 silicon is being developed now, and initial 3.1 cable modems should be ready for trials at the end of this year and the beginning of next year. Headend equipment trials and early deployments will likely follow in 2016.
Gerry White, distinguished engineer at Cisco Systems Inc. (Nasdaq: CSCO), said one of the big questions for vendors to answer is "Where is the starting point we want to go for?" In other words, how much change should vendors try to accommodate in the first round of DOCSIS 3.1 products, and how much should be implemented in future phases? In general, he said, there has to be enough power, cooling, and interconnect bandwidth from day one, but many other elements in the hardware will simply be upgradeable as operators need additional capacity or features.
Ultimately, the actual rollout of DOCSIS 3.1 equipment will come down to operators' specific network needs. For cable modems, a big determining factor for adoption in the near future will be the price delta between DOCSIS 3.0 and DOCSIS 3.1 products. As Salinger said, because DOCSIS 3.1 modems will initially be deployed in DOCSIS 3.0 mode until new headend equipment is installed, it will be difficult to justify spending a lot more money. However, if the price difference is minimal, then the decision to upgrade will be much easier to make. (See Docsis 3.1 Stays on a Fast Track.)
Marty Davidson, vice president of engineering and network operations at Society of Cable Telecommunications Engineers (SCTE) , said operators should start thinking today about new DOCSIS 3.1 training programs. Even though the technology is still under development, it makes sense to begin educating engineers and even technicians about what DOCSIS 3.1 will enable.
At this time next year, Salinger said, he believes that the industry should be ready to talk about DOCSIS 3.1 interop events. That may be an optimistic statement, but there's no doubt that industry technologists expect a lot to be accomplished with next-generation cable architectures in a very short period.
And DOCSIS 3.1 is only one piece of the puzzle.
— Mari Silbey, special to Light Reading