The Bauminator

Tools of the Trade

Cable Still Has Plenty of Ways to Stay Ahead of Bandwidth Curve 3/21/2016 7:15 AM

It’s a question that has been posed for as many years as I’ve been covering this industry: When will cable operators take fiber all the way to the home?

 

The answer has always been something like: Someday, but not any day soon.

 

And I don’t see that answer changing much after listening to Tom Cloonan, the chief technology officer of Arris’s Cloud & Networks unit, speak last week at the company’s Investor Day.

 

Cloonan has been tracking bandwidth trends for since “broadband” entered the lexicon, having joined Arris in 2002 via its acquisition of Cadent, then a pioneer in the world of DOCSIS-powered network equipment.  His track record in this area, and many others, has been spot on as the cable industry continues to stay ahead of the voracious demand for bandwidth, which continues to grow at an annual rate of about 50%.

 

Sure, cable operators are using fiber-to-the-premises technologies today, but in a limited form. Comcast, for example, uses FTTP for Gigabit Pro, its new symmetrical 2 Gigabit per second residential service, but that is being targeted only to a sliver of the market that actually wants or needs such a  service.  Cox Communications is also using it for the early phase of its 1-Gig residential service deployment. MSOs are also going all-fiber in most greenfield/new-build scenarios.

 

But, again, these are limited, targeted deployments.  Thanks to technologies like DOCSIS 3.0, widely deployed today, and DOCSIS 3.1, cable’s new multi-gigabit platform, there still’s plenty of gas left in the tank of the hybrid fiber coax network.

 

Back to Cloonan. He readily admits that bandwidth demands have never been greater, thanks to the popularity of over-the-top video, the emergence of higher-resolution 4K video, the Internet of  Things, and the coming wave of virtual reality.  

 

Is the industry entering  time of inflection? he asked. “I would have to say the answer is definitely yes,” Cloonan said.

 

And the problem (or challenge, if you’re an optimist like Cloonan) was illustrated  in a chart that showed the expected demands for bandwidth in the years ahead. It clearly indicates that today’s HFC plant (most is built to 860 MHz) will soon be overwhelmed, perhaps as soon as 2018 … if cable operators were to sit on their hands and do nothing.

 

High-speed data over DOCSIS, Cloonan said, “is a monster.”  Also on the horizon is a “simulcast bubble” that will apply significant bandwidth pressure as MSOs deliver their pay TV service simultaneously in QAM and in IP during a transition that will likely span many years.

 

While the obvious answer to this dilemma is to boost the spectrum range, there are many other tools in the toolbox to keep cable ahead of the curve before it comes to that.

 

Among them are DOCSIS 3.1, which is expected to be 50% more bandwidth-efficient than DOCSIS 3.0, new distributed access architectures, new codecs like H.265/HEVC, the possibility of a proposed “Full Duplex” technique that could enable cable to offer symmetrical speeds on DOCSIS, higher density equipment, and new “virtual” architectures.

 

And, longer term, it’s possible that MSOs could raise the ceiling well above 860 MHz or even 1 GHz. One of the proposals Arris is working on is pushing spectrum to 2 GHz or even 6 GHz or higher – enough to enable DOCSIS plant to support capacities of 50 Gbps or more.

 

But make no mistake that cable will pull fiber closer and closer to the home. Cloonan also believes there’s a major inflection point coming with respect to node splits, which increase the amount of bandwidth that’s delivered to smaller and smaller groups of homes.

 

Looking into the relative near-term future, Cloonan sees a 10x increase in nodes. For example, if Arris’s flagship converged cable access platform, the E6000, serves about 112 nodes today, tomorrow’s more distributed architectures could push that to about 10,000 nodes.

 

If all of these tools and options turn out to be technically feasible (while also being cost-efficient), it would seem that an en masse shift to FTTP by MSOs is still very far out on the horizon.