CMAP: Next-Gen Platform Takes Aim at Convergence
The ultra-dense box will save on space and powering while providing more capacity.
For cable operators, the waves of narrowcast services are only going to get bigger and come crashing in with more frequency in the near future. In order to stem the tide, Comcast has been spearheading its Converged Multiservice Access Platform (CMAP) project with vendors, cable operators and other cable industry leaders.
Jorge Salinger, Comcast’s vice president of access architecture, has been on the tip of the spearhead for Comcast in regard to developing an ultra-dense platform that not only meets the demands for managing more video-on-demand, network DVR, switched digital video, and DOCSIS services and implementations, but also lays the groundwork for a passive optical network for business or greenfield deployments, wireless technologies, IP video and future access technologies.
With cable operators reducing the size of their service groups as a backdrop, the high-level goal of CMAP is to collapse the cable modem termination system and edge QAMs into a platform that also takes up less space and uses less power than the current technologies.
“The objective is to have an architecture that vendors can implement that allows for much more density and significant power savings and operational simplifications,” Salinger said. “The main reason is the explosive growth of narrowcast services. Fortunately, they are very successful services, and because of that we need to support the corresponding growth of the deployment of QAMs. Generally speaking, two things can happen: The boxes get denser or you need more space. The boxes are getting denser, but not at the rate that would allow us to keep the headends at the same space with the same powering.
“The savings going from the current implementation to the CMAP implementation are probably 50 percent space savings in the same headend and more than 50 percent in power savings.”
Salinger said Comcast typically has about 10 narrowcast QAMs in a headend – six for VOD and four for HFC – but with CMAP, the same number of services could be provisioned in half the amount of space with four times the number of QAMs than the current implementation.
“The CMAP architecture offers efficiencies that help us be mindful of power consumption needs and space consumption needs while achieving technological innovations that are aligned with the density, or cost-per-port, type of economics,” said Matt Bell, Charter’s vice president of engineering and development. “Charter is always interested in those aspects so we can provide the best products for our consumers.
“One of the parts of the formula that we’re always turning the knob on is in regard to service group size and the density. The ability for us to stack more QAMs in a port allows us to get very precise with service group sizing, which is a great benefit for customers with regard to capacity and congestion.”
CMAP is a keystone for Comcast’s Next Generation Access Architecture (NGAA) initiative, but while Comcast is driving the effort, it has picked up other passengers, including Cox Communications, Cablevision, Rogers Communications, Charter Communications, Liberty Global, European cable operators at large led by Cable Europe Labs and the National Cable Television Cooperative (NCTC). While it hasn’t signed an agreement to date, Time Warner Cable is also providing its input on CMAP.
CMAP can be implemented in two ways: integrated with all of the functions in one chassis, or modular with the CMAP functions divided between a Packet Shelf (PS) and Access Shelf (AS). Packet Shelf implements the packet processing functions, such as subscriber management, service flow management, video program stream edge manipulation and Layer 3 routing.
The AS implements all of the upstream and downstream PHY functions that are normally associated with CMTSs and edge QAMs and as much of the MAC that is needed to support both upstream and downstream flows. An interface between the AS and PS is needed to enable interoperability between AS and PS vendors, according to a white paper by Salinger.
By dividing CMAP into integrated and modular versions, cable operators have more flexibility for which implementation they choose, and more vendors can take part in the process.
“There’s no one solution fits all,” said Gil Katz, Harmonic’s senior director of cable solutions. “I think there’s room for modular and room for integrated approaches. It all depends on the size of the headend, or in some cases the size of the hub/OTN architecture. We think the fact that CMAP supports both architectures is the right thing to do.”
Katz said Harmonic wasn’t ready to share its CMAP product strategy, but the company is a provider of edge QAMs to Comcast, so the modular approach may be in the offing. Other vendors that are actively pursuing CMAP-related products include Motorola, RGB Networks, Arris, LiquidxStream Systems and BigBand Networks. During a summer earnings call, Arris provided a glimpse of its CMAP product when CEO Bob Stanzione mentioned the company’s “E6,” which he described as a combination of an edge QAM and CMTS.
With Packet Shelf, router companies such as Juniper Networks, Alcatel-Lucent and Huawei have the opportunity to develop CMAP-related products, while edge QAM vendors can work on Access Shelf products without building full-fledged routers.
“It allows vendors that don’t have the expertise to build the whole thing to stay in the business, and it allows us to have best-in-breed, as well,” Salinger said of the modular/integrated approach.
Comcast first conceived the CMAP concept in 2008, followed by the start of specification development last year. Working with vendors, cable operators, and advisors such as CableLabs and Cable Europe Labs, specification development was divided up into three phases. In March of this year, the first version of the hardware and functions specification was completed, followed by a revision in July. Also in July, the first versions of the configuration and management and modular interface specifications were completed. The second and final iterations of the modular access shelf-to-packet shelf interface (PASI) and configuration and management specifications, along with the third version of hardware and functions, will be wrapped up in November.
Once the specs are finished, Salinger said testing would start next year, followed by more product availability in 2012 and deployments in early 2013.
FLEXIBILITY KEY TO CMAP
Whether it’s modular or integrated, Arris chief strategy officer Tom Cloonan said CMAP will be deployed over the next decade in different flavors based on cable operators’ architectures and needs. While Time Warner Cable and Charter have deployed a large number of narrowcast QAMs for SDV, Comcast is using a large number of digital QAMs for its analog-to-digital bandwidth reclamation project.
“At a high level, CMAP offers a lot of potential benefits for the MSOs, but having the flexibility to multiplex all of those services through a single RF connector is a key parameter,” Cloonan said. “Comcast has been pretty vocal about it, and Cox has been pretty vocal about it. I think a lot of the other MSOs are still wrapping their arms around what CMAP is from a definitional point of view. I think it matches a lot of what other MSOs are asking for, even though they may not put the label of ‘CMAP’ on their description yet.
“Overall, I think most MSOs are seeing the future evolve in this fashion, where several needs are going to be required of their nextgeneration boxes, and those needs include, probably first and foremost, much higher capacity in terms of channels per service group than they were supporting in the past. Generically speaking, they’re also wanting to simplify the overall management and configuration of the box, whether you’re turning on or off the channels, changing the multiplexing of channels or enabling the frequency of the channels.”
Michael Cookish, a senior director of CMTS product management at Motorola, said that CMAP is drawing a lot of interest abroad, particularly in Europe where space is even more limited for the expansion of headends.
“Along with collapsing CMTSs and edge QAMs, European cable operators are also some of the biggest consumers of power,” Cookish said. “It’s a big deal for these operators to add more channels while being able to manage and lower their energy usage.”
As for smaller Tier 2 or Tier 3 cable operators, or even smaller headends owned by the large MSOs, Harmonic’s Katz said that since CMAP is all about port density, it will scale down to meet the needs of smaller headends.
“As long as it maintains a certain density per port, it can be four rack units, six rack units or 10 rack units,” Katz said. “By the time CMAP is ready, a Tier 2 cable operator, by segmenting to 300 homes in a service group, can serve about 300,000 subscribers. Or you could use a modular approach if you have connectivity between different sites.”
WHAT CMAP ISN’T
Comcast’s Salinger is quick to point out that CMAP isn’t a new specification for DOCSIS or any other CableLabs specification. It does draw on CableLabs specifications, such as DOCSIS RF Interface and the DRFI specifications, to create the product specifications.
While CMAP is designed to take advantage of the current specifications, it will also evolve with the industry, which at some point will include new silicon from vendors, more wireless technologies – if desired – and advanced access technologies.
“I think the MSOs wanted, Comcast and Cox in particular, but also other MSOs, as well, to be able to port different PHYs in the future and make sure the CMAP chassis can support new PHYs if and when CableLabs develops some new PHY tools that will allow more upstream bandwidth,” Arris’ Cloonan said. “There are ongoing discussions on those types of topics within the MSOs, and they want the CMAP spec to support new upcoming PHYs in the future.
“I think there has been a good joint effort between Comcast and the vendors. The vendors wrote a lot of the spec, and I think it’s a good spec. It’s defining an exciting future for the whole industry. It’s going to be big, and I think it’s the right thing for the industry to do because of the flexibility it offers. We’re anxious to get there.”