Does Docsis Have a 10-Gig Future?

BOSTON -- The Cable Show -- Engineers from Arris Group Inc. (Nasdaq: ARRS), Cisco Systems Inc. (Nasdaq: CSCO) and Motorola Mobility LLC have joined forces to refresh cable's Docsis platform and put it on a path that, they claim, could achieve a capacity of 10Gbit/s in the downstream and 2Gbit/s in the upstream.

That would blow the latest generation of Docsis 3.0 technology out of the water. Here at the show, Intel Corp. (Nasdaq: INTC) is previewing its new Puma6 family of chips, including a "media gateway" configuration that bonds 24 downstream channels and eight upstream channels -- enough to produce speed bursts of almost 1Gbit/s downstream and 320Mbit/s upstream. (See Intel's New Docsis 3.0 Chip Guns for 1-Gig .)

John Chapman, the CTO of Cisco's cable access business unit, presented the ideas here Tuesday, offering a condensed version of a 182-page paper co-authored by the vendors, which compete in the market for Docsis cable modems and cable modem termination systems (CMTSs).

The paper's stated aim was to give the cable industry recommendations on how to develop a new generation of Docsis as cable operators start to migrate all of their services to IP.

But there's a bit more to the agenda, as the vendors also have a vested interest in protecting their hold on the Docsis equipment market as other potential technical options begin to emerge, particularly EPON Protocol Over Coax (EPoC), an Institute of Electrical and Electronics Engineers Inc. (IEEE) standard in the works that's looking to bring PON-like performance to hybrid fiber/coax (HFC) networks. (See The Docsis Addendum and EPON-Over-Coax Starts Its Standards Journey .)

The Docsis vendors listed out 10 areas of consensus that, they believe, could start to redefine Docsis and create a roadmap that would enable the platform to achieve the kind of speed performance that EPoC is also shooting for. Here are a few ideas in which Arris, Cisco and Motorola are in agreement:

  • Define backwards compatibility as a goal that will let current Docsis cable modems and next-generation Docsis modems use the same spectrum.
  • To start beefing up cable's skinny upstream path using a "mid-split," a process that would expand cable's usable upstream spectrum from today's range of 5MHz to 42MHz, to 5MHz to 85MHz. Longer-term, the new specs would also define a "high-split" that would expand the upstream even further -- perhaps up to 400MHz.
  • On the downstream side, the vendors suggest operators manage that spectrum using tools like switched digital video (SDV) and H.264 and upgrade to 1GHz. Longer-term, they suggest operators raise the ceiling to 1.7GHz.
  • Move to higher orders of modulation (4096 QAM or more) that will increase the number of bits that can be transmitted on the cable plant.

Chapman didn't take any direct shots at EPoC, but did call Docsis "the most successful Ethernet-over-coax technology to date."

And he argued that Docsis can be molded into something that can handle cable's future performance requirements, so long as the industry agrees to commit to a new roadmap and set to work on it soon. "Docsis can be anything the Docsis community wants or needs it to be," Chapman concluded.

— Jeff Baumgartner, Site Editor, Light Reading Cable

joanengebretson 12/5/2012 | 5:32:21 PM
re: Does Docsis Have a 10-Gig Future?

It seems like there is a lot more room to increase the bandwidth that hybrid fiber coax networks can support. In contrast, I'm hearing that DSL is largely tapped out. 

Telcos will likely need to go to fiber-to-the-home to really compete effectively with cable. But they still seem to have trouble justifying that investment from a revenue perspective outside areas where FTTH is already deployed.

No wonder Verizon Wireless is partnering with the cable companies.


Jeff Baumgartner 12/5/2012 | 5:32:07 PM
re: Does Docsis Have a 10-Gig Future?

There does appear to be alot of gas in the HFC tank before taking that final FTTP step. For a while it seemed that Docsis and EPoC could grow up together, but it's starting to appear that cable might face a fork in the road at some point. Does it make sense to remake Docsis (4.0?) with their current set of vendors or try to expand further into the Ethernet world and its deeper pool of suppliers by going with EPoC?  It'll make for an interesting technology (and political) debate , but probably too early to say which way it will sway. JB

joanengebretson 12/5/2012 | 5:32:06 PM
re: Does Docsis Have a 10-Gig Future?

It also may not be an either/or situation -- at least not for some years. Fiber-based cable solutions will probably be deployed first where the economics make the most sense, just as we saw with PON, while in other areas network operators will try to work off of existing infrastructure for as long as they can.

Duh! 12/5/2012 | 5:32:05 PM
re: Does Docsis Have a 10-Gig Future?

It will be interesting to see all this play out. Some of these architectures will be more widely deployed than others, some might even be permitted to die.  But I don't expect to see a discrete "fork in the road", for several reasons.   Anybody who's worked with MSOs understands that getting any kind of big picture architecture consensus is like herding cats.  In addition to the tendency of colorful personalities with strong opinions to thrive in MSO corporate cultures, there is too much diversity in the installed base, deployment scenarios, population density, competitive environment and local regulation to get to one solution any time soon.  There can be no doubt that MSOs will pursue the lowest cost solution, but how they perceive the balance of CAPEX and OPEX will vary, along with the many contributing factors to each.   I also would not underestimate the power of relationship sales in keeping the incumbent vendors incumbent. 



Sign In