NEW ORLEANS – With the Gigabit Era clearly upon us, cable operators are now fervently searching for the best way to enter that era and the best next-gen technologies to make that entry successful.
This search was very much in evidence at last week's SCTE Cable-Tec Expo show in New Orleans. Over the course of three days, teams of MSO technologists marched from booth to booth on the exhibit floor, hoping to uncover the right mix of equipment and software that would enable them to introduce gigabit speeds sooner rather than later (here in Canada, Rogers and Eastlink – not to mention Telus and Bell, have made Gb announcements). MSO engineers kicked the tires on a wide range of equipment – cable modems, wireless broadband gateways, cable modem termination systems (CMTSs), Converged Cable Access Platform (CCAP) devices, testing products and more – in their quest for the magic formula.
Not surprisingly, the vendors on the exhibit floor were more than happy to oblige the operators, serving up lots of fresh DOCSIS 3.1 and CCAP gear to entice them. In particular, such major equipment suppliers as Arris and Hitron Technologies took advantage of the show to demonstrate their brand-new DOCSIS 3.1 modems, which are capable of supporting broadband speeds as high as 5 Gbps in the downstream and 1 Gbps to 2 Gbps on the upstream end.
At the same time, Arris and such other big equipment suppliers as Casa Systems, Cisco Systems, Gainspeed and Huawei Technologies demonstrated their versions of integrated CCAP chasses and/or distributed CCAP architectures. Vendors are developing these denser, more efficient systems, which combine the data-processing functions of the cable modem termination system (CMTS) and the video processing functions of the edge QAM modulator in the cable headend, to support the greater bandwidth capabilities of the new DOCSIS 3.1 spec.
Although it's doubtful that any big equipment purchase decisions were made in the Big Easy, an industry consensus for rolling out both DOCSIS 3.1 and CCAP does seem to be developing. In spite of the headline-grabbing D3.1 rollout plans recently unveiled by such large North American and global MSOs as Comcast, Rogers and Liberty Global and such smaller ones as Cogeco-owned Atlantic Broadband and General Communication Inc. (GCI), it appears that most cable operators will introduce both gigabit-enabling technologies gradually in a phased approach over the rest of the decade, not all at once.
For example, in an informal poll conducted during a morning panel session on DOCSIS 3.1 last Thursday, most cable operators in the room indicated that they do not plan to roll out D3.1 across their regions next year. Instead, most said they would start D3.1 rollouts in 2017 or possibly even later. Thus, DOCSIS 3.0, and even DOCSIS 2.0, will not exactly disappear from the cable landscape any time soon.
In a sign of this cautious, incremental approach, the first group of DOCSIS 3.1 modems hitting the market will be hybrid models capable of supporting both the D3.1 and D3.0 specs. Expected to cost about 30% to 50% more than standard D3.0 modems initially, these new hybrid models will enable cable operators to bond 32 downstream and 8 upstream channels under the D3.0 spec, as well as use blocks of subcarriers based on orthogonal frequency-division multiplexing (OFDM) technology to support even higher speeds under the new D3.1 spec.
"If you do [upgrade], the benefits that DOCSIS 3.1 brings are far more than anything before.” – Jorge Salinger, Comcast
As a result, the new modems will offer MSOs the best of both worlds, according to leading cable technologists. "DOCSIS 3.1 modems will operate like the best DOCSIS 3.0 modems you can buy," asserted Jorge Salinger, VP of access architecture at Comcast. Speaking on the panel last Thursday, he said cable operators will be able to deploy the new modems without making any upgrades to their networks or headends.
But, Salinger noted, if cable providers do upgrade their systems for DOCSIS 3.1, they will be able to reap a slew of benefits, including much faster broadband speeds, more bandwidth for IP video and other new IP-based services and greater operational efficiencies. "If you do [upgrade], the benefits that DOCSIS 3.1 brings are far more than anything before," he said.
It also seems clear that as they make the move to D3.1, cable operators will roll out CCAP to support the surging bandwidth demands of gigabit services, IP video, Ultra HD (4K) video, WiFi and the like. That's because their existing combinations of CMTS and edgeQAM devices will likely not be up to the challenge of handling all those bandwidth-hungry services. In fact, cable operators have already been making the move to CCAP in advance of DOCSIS 3.1, as shown by contract orders for the major equipment vendors.
The big question with CCAP, though, is which type of architecture the MSOs will choose to deploy. Indeed, the debate is already heating up over the virtues of the various versions of CCAP architecture that are now or will soon be available, including the integrated CCAP chassis model that's already out in the market and several new distributed CCAP architectures that are now being developed.
All three of the main distributed CCAP architectures call for moving at least part of the CCAP functions and equipment out of the cable headend and into the access network at the fiber-optic node level. But the three versions differ in exactly how much of the functions and equipment they would shift out of the headend and into the access network.
At the SCTE show, the four big distributed CCAP players – Casa Systems, Cisco, Gainspeed and Huawei – all made the case for their particular flavor of the technology. But the cable industry is far from developing a consensus on the best path to take just yet. In a technical workshop on the subject last Thursday, for example, Tom Cloonan, CTO of Network Solutions for Arris, observed that all three options have their pros and cons.
"What's best?" Cloonan asked. "I don't have the answer." Punting on the question, he said he thinks that "all of them are workable solutions and technically feasible."
