& cplSiteName &

Here Comes DOCSIS 4.0

Alan Breznick

AUSTIN, Texas -- Big Communications Event -- If you thought that Full Duplex DOCSIS might be the last of the DOCSIS specs, you're in for a surprise. Even as the cable industry rolls out DOCSIS 3.1 en masse and preps for the next-gen Full Duplex spec now in the works, some leading cable technologists are already looking at what comes next.

Speaking during a fireside chat and in a later interview here last week, Cox Communications Inc. 's executive director of advanced technology, Jeff Finkelstein, revealed plans to craft a futuristic DOCSIS spec for the next decade that would go well beyond DOCSIS 3.1 and Full Duplex to enable symmetrical broadband speeds as high as 30 gigabits per second or even 60 Gbit/s. DOCSIS 3.1 can enable speeds as high as 10 Gbit/s downstream and 2 Gbit/s upstream, while Full Duplex is designed to enable symmetrical speeds as high as 10 Gbit/s. (See Cable Closes In on 5-Gig Upstream With FDX .)

Referring to the proposed spec as "DOCSIS 4.0" or "DOCSIS dot Next," Finkelstein explained that it would support such high speeds by tapping into more of the cable HFC plant's available RF spectrum. Under this idea, also known as "Extended Spectrum DOCSIS," cable operators could use more than the 1.2GHz of plant spectrum that can now be leveraged for DOCSIS 3.1 or the 1.8GHz of spectrum envisioned for Full Duplex. (Note that plenty of operators aren't even using 1GHz yet, but the vision is there.)

Specifically, Finkelstein contemplates using at least 3GHz of spectrum for the new spec, nearly tripling the amount of bandwidth available, and potentially as much as 6GHz. (The coaxial cable physical limit is around 10GHz.)

"Think about it," he says, noting that cable keeps boosting the amount of RF spectrum it uses. "Why can't we go to 3GHz?"

Finkelstein is not the first cable technologist to propose an extended spectrum version of DOCSIS publicly. Back in May 2016, for instance, Tom Cloonan, CTO of network solutions for Arris Group Inc. (Nasdaq: ARRS), advanced the concept at the last big NCTA convention in Boston. At that time, Cloonan called for using 6GHz of spectrum and Fiber Deep to deliver speeds of 50 Gbit/s or more.

But, building on Cloonan's concept, Finkelstein has run the calculations with a team of Georgia Tech graduate students and found that the physics behind Extended Spectrum DOCSIS could actually work. Consequently, he plans to start promoting the idea with CableLabs and key industry vendors like Arris and Cisco Systems Inc. (Nasdaq: CSCO), with the goal of crafting a full spec by the mid to late 2020s.

For more of Jeff Finkelstein’s ideas, listen in to a radio show on our sister site, Broadband World News Thursday, May 24 at 1 p.m. ET.

So how would Extended Spectrum DOCSIS work? Finkelstein proposes using advanced wave form technology instead of the orthogonal frequency division multiplexing (OFDM) modulation scheme currently employed for DOCSIS 3.1.

"I'd love to really start digging into this," he says. "This is the time to start moving the ball forward."

Finkelstein notes that Cox ran data tests with 3GHz spectrum as far back as 2005 and proved that the extended spectrum approach could work in the lab. Now, he says, it's a question of whether it can work economically in the field.

Of course, as Finkelstein acknowledges, extending the RF spectrum further for DOCSIS is no simple matter. "The physics are hard," he says. "And the business case still needs to be rationalized."

But, noting that DOCSIS 3.1 seemed like a "crazy" idea when he and others first proposed it back in 2007, he feels confident enough about the extended spectrum concept to go public now. "We can make magic happen," he says. "We proved that with DOCSIS 3.0 and other enabling technologies."

Could be another reason why DOCSIS will just keep going on. (See Why DOCSIS May Never Die.)

— Alan Breznick, Cable/Video Practice Leader, Light Reading

(4)  | 
Comment  | 
Print  | 
Threaded  |  Newest First  |  Oldest First        ADD A COMMENT
Kelsey Ziser
Kelsey Ziser,
User Rank: Blogger
5/22/2018 | 4:05:35 PM
DOCSIS dot Next
What's after DOCSIS dot Next? DOCSIS dot Next-Gen? 
Clifton K Morris
Clifton K Morris,
User Rank: Light Sabre
5/23/2018 | 5:10:01 PM
Re: DOCSIS dot Next
Not sure what they call it, but perhaps a rebranding is in order... I know people who have moved for FTTH service, only to switch to Cable within 6 months. It seems the FTTH providers lack diversity in their peering agreements.

That’s definitely a strength that could be played up in cable marketing efforts, and in a world where “up-to” speeds are the competition’s marketing norm.
User Rank: Moderator
5/23/2018 | 2:08:01 AM
Clearly after DOCSIS 3.1, we need...
DOCSIS for Workgroups 3.11, then...

DOCSIS 95, then...

DOCSIS NT 4, then...

DOCSIS 98, then...

DOCSIS ME, then...

DOCSIS XP, then...

DOCSIS Vista, then...

DOCSIS 7, then...

DOCSIS 8, then...

DOCSIS 8.1, then...


And live updates every so often after that where there's a 25% chance to brick the customer's modem.
User Rank: Lightning
6/27/2018 | 2:19:41 PM
Fiber-cable could push further
Signal modulation schemes used in fiber optic and cable links have often led technology due to the controlled environment that the transport media provides.  OFDM, for example, was used in fiber-optic several years before it was made part of the mobile wireless standards.  Fiber and cable also benefit from being line-powered rather than battery powered, which allowed early adoption than use in mainstream mobile devices.  

Some advanced signaling methods require that the signaling conditions are made stable and more optimal than can feasibly be achieved at this juncture using wireless as the media.  Among the most adventurous of these methods is the use of non-sinusoidal waveforms such as fractal waveforms.  That method uses the waveform itself as a data encoding mechanism: instead of being a one or zero, the fractal-encoded signal stream would be a massively-oversampled fractal pattern that represents a code or 'word'.  Groups of such words would be transmitted using error correction validation.  That is analogous to using a highly accurate form of short-hand dictation: many times more data would be encoded than today's sinusoidal signaling methods. This example may never happen. However, it exposes why fiber and cable might evolve to use signaling methods that take advantage of the controlled media environment that could go well beyond DOCSIS 4.0/dot next as it is most commonly now considered. 



Featured Video
Flash Poll
Upcoming Live Events
September 17-19, 2019, Dallas, Texas
October 1-2, 2019, New Orleans, Louisiana
October 2-22, 2019, Los Angeles, CA
October 10, 2019, New York, New York
November 5, 2019, London, England
November 7, 2019, London, UK
December 3-5, 2019, Vienna, Austria
December 3, 2019, New York, New York
All Upcoming Live Events