The cable industry is getting ready to unleash PacketCable Multimedia (PCMM) on an unsuspecting world. Though it's still not entirely predictable what will happen when operators let PCMM out of its cage, they will surely be able to harness some of that pent-up energy to introduce innovative new features and applications, and boldly enter new markets. But not yet. Probably in 2007. And because PCMM is, in essence, an enabling technology, you (or cable's customers, at least) might not even notice.
![]() |
PCMM goes beyond "best effort" to assure quality of service (QoS), making it possible for operators to dynamically allocate the network resources appropriate for a range of IP-based data, video, and voice applications.
Very few companies acknowledge using PCMM today, and one of the most prominent applications widely assumed to be based on PCMM isn't.
Few operators will discuss their plans to use PCMM, because those plans constitute strategic information, an overused excuse which, in the case of PCMM, is probably justifiable.
PacketCable Multimedia is an extension of PacketCable. PacketCable is a set of specifications that leverage DOCSIS 1.1 and can deliver IP-based voice service on a two-way cable plant.
The first PacketCable specification, PacketCable 1.x, describes a way for cable operators to install and run equipment that mimics a telephone switch, providing the basis of voice-over-IP (VoIP) telephony service. The PacketCable 1.x architecture also allows operators to manage their bandwidth in such a manner that VoIP service can be delivered reliably.
![]() |
Figure 1: As defined by PCMM specs,
policy servers will enforce traffic rules for a variety of network applications. |
PacketCable Multimedia makes the PacketCable 1.x bandwidth management architecture generic by enabling support for almost any application or service based on Internet Protocol. PCMM defines an architecture where the operator can introduce a new application with the installation of an application manager. Application managers request network resources on behalf of a client; policy servers authorize and commit those requests. PCMM can also support peer-to-peer and client-based applications as well.
PCMM is not PacketCable's finale. PacketCable 2.0 has been developed to facilitate the complete convergence of voice, data, video, and mobility services on a DOCSIS network with a modular architecture. Some of the functions being defined by PacketCable 2.0, according to CableLabs, include the capability for enhanced residential VoIP services and video telephony, feature integration across service platforms, and the ability to track and transfer applications across a variety of client devices across service platforms.
Among the easiest things to do with PCMM is improve existing services. An operator can assure its customers that voice or video quality will improve, or that players of online games will not be plagued with annoying lag, for example.
PCMM allows a service provider to, for instance, increase and decrease the amount of bandwidth required by any given subscriber for a given application. If the operator detects that a particular subscriber has purchased a film and is about to download it, the operator could, for example, increase the data transmission rate to that subscriber for the duration of the download to make the download go faster.
An operator could also prioritize packets. If a subscriber is downloading a movie while playing an online videogame, an operator can make sure game data is routed promptly.
Service quality could easily be offered for free, but some suggest that operators could, for example, offer QoS to their online gamers for a fee, and call it a Game Tier.
Because video and online games are two applications most susceptible to latency, delay, and jitter, they are the two applications expected to benefit the most from the introduction of PCMM. Of the very few who have deployed PCMM, however, only one has a game application, and none are doing video.
"People are still looking at its applicability to online gaming, and also to business service applications, both voice and data to small and medium businesses," says Glenn Russell, CableLabs' director of multimedia applications.
StarHub last year contracted with Motorola to upgrade its system to PCMM. The Singapore-based operator is using PCMM to 1) create a "turbo" button that allows subscribers to order a burst of bandwidth, 2) facilitate Xbox online gaming traffic, and 3) to enable SIP-based VoIP.
StarHub's upgrade included Motorola's PCMM-compliant Broadband Services Router (BSR) 64000 cable modem termination system, Sandvine's PCMM application manager, Camiant Inc.'s policy server, and provisioning software from both Incognito and Interactive Enterprise.
In January, Buckeye CableSystem in Toledo, Ohio, said it would use PCMM to provide its customers with the ability to test drive a higher-bandwidth data tier.
Buckeye CableSystem deployed Camiant's PCMM-compliant QBUS policy server and Camiant's Bandwidth on Demand Application Manager.
Buckeye has four data speed tiers: 8 Mbps/768 kbps, 5 Mbps/512 kbps, 1.5 Mbps/128 kbps, and 96 kbps symmetrical. Operators can let subscribers test a faster tier without having to send any reconfiguration files, and without subs having to reboot their cable modems.
![]() |
Figure 2: Network elements defined by PCMM will pave the way for IMS integration.
|
Buckeye CableSystem will also enable temporary speed increases for specific applications. Subscribers moving large files can click a Turbo button to get a burst of bandwidth.
Buckeye CableSystem plans to eventually offer SIP-based telephony and other IMS (IP Multimedia Subsystem) applications using PCMM, but has set no schedule for introducing those services.
Cox Communications, meanwhile, is also deploying PCMM policy servers from Camiant, and is using the system to give cable modem subscribers temporary trials of higher-speed service tiers. Cox has been offering speed previews for years. The difference is that without PCMM, Cox has to batch process groups of modems, sending reconfig files to set up the preview, then sending more reconfig files once the trial period was over to reset subscribers' modems. Now with PCMM, Cox can provide the feature on a per-subscriber basis, if it so desires.
Cox would not discuss specific numbers, but says the number of subscribers upgrading their service has in many cases exceeded expectations. "Our customers are very satisfied," says Scott Hightower, VP of data product development at Cox. "We're pleased with the return on investment. This has more than paid back the investment on the platform."
Cox is considering other applications that can leverage PCMM technology, but Hightower says the company has yet to identify the next one it wants to deploy–which was the company's position in May when it first announced its deployment of PCMM.
The technology will definitely be leveraged, however, he says. Both the residential and commercial business sides of Cox are exploring what they can do with PCMM.
Comcast in September began offering "PowerBoost," an app that automatically and temporarily accelerates transmission speeds for customers transferring large files, provided the bandwidth is available. Because the feature is similar to Buckeye CableSystem's, there was some expectation that it, too, is a PCMM-enabled application. It is not. The capability was created with a proprietary, patent-pending bit of software added to Comcast's cable modem termination systems. The CMTS automatically provides the bandwidth boost whenever excess is available.
Operators do not have to roll out PCMM to offer such capabilities, it's clear. But as always, it's a matter of tradeoffs. Comcast was able to implement the capability across its entire vast footprint with a simple software upgrade, whereas Cox has to add equipment in every system in which it wants to provide speed previews.
On the other hand, once Cox or any other operator has installed the PCMM platform in its systems, the introduction of new applications will be a fairly simple matter of adding application servers.
Neither approach is necessarily better than the other. The point is that even without PCMM, operators have the flexibility to perform bandwidth tricks, and that may be one reason why PCMM isn't as widely implemented as we approach the end of the year as many expected at the beginning of 2006.
Camiant is participating with all three publicly-announced PCMM implementations, and professes to be involved with a great many tests and trials with other companies. The company says two of the key drivers for PCMM are SIP-based voice and video.
If SIP-enabled VoIP is a key driver, that may be another reason why PCMM hasn't inspired headlines in 2006, StarHub in Singapore notwithstanding. U.S. operators are likely to deploy a SIP-based telephony service based on standards, and the SIP telephony spec for PacketCable 2.0 was published only at the end of September.
And part of it may be that some operators are just keeping mum.
"Every major operator is talking about policy platforms, if not for '06, then for '07," says Ed Delaney, Camiant's vice president of marketing. "They're all experimenting with it or quietly using it."