CableLabs called it a "major milestone." The qualification of six PacketCable Multimedia (PCMM) devices in late April was indeed noteworthy. Reaching this point could have taken more than one certification wave (CW). As it was, CW 34 saw the qualification of PCMM policy servers from all five of these vendors: CableMatrix, Camiant, C-COR, Tazz Networks and Telcordia. Additionally, CableLabs qualified Motorola’s BSR 64000 cable modem termination system (CMTS) as PCMM-compliant. Anticipating activity on the PCMM front, Communications Technology hosted a roundtable discussion earlier this year. "Virtually every operator has had a trial of PCMM in 2004," said Camiant Vice President of Business Development Edward Delaney. "2005 is about deployments." To gain greater understanding of PCMM and its potential, see the following edited transcription of the conversation that CT held with Delaney and Senior Manager, Cisco Systems, Chris Maloney; Director of Advanced Network Technologies, Charter Communications, Michael Emmendorfer; and VP Broadband Technology Integration, Cablevision Systems, Joe Godas. CT: Let’s start off with a positive. What do you each of you see as PCMM’s strongest potential for cable? Delaney: I think that the opportunity to have a variety of premium applications, all prioritized with quality of service over the high-speed data network, is its greatest potential for cable. Maloney: The reason that PCMM is so powerful and has so much potential is because it allows you to intelligently use the bandwidth in ways that you never could before. It allows you on a per-application basis to define what you’re going to do and what you’re going to set in terms of policy and QoS, especially over the HFC plant, even when you don’t know in advance what the applications are going to be. That flexibility is something the operators have never really had before. They’ve always had to plan what applications they’re going to have nine months, a year, two years, three years out in their network, and now they don’t have to. Now they can say, "I need more bandwidth for this application, and I want to turn it on tomorrow." Emmendorfer: One of the main purposes of PCMM is going to be revenue generation and service creation at a high level. Those are some of the key business drivers. Much of the things that Chris identified in terms of the technical pieces that can ensure serviceability and service assurance, as well as service creation from PCMM, can be implemented. Some of the QoS features we can implement using CMTS technology, as well as other technologies that will emerge over the next few years. We may be able to leverage other services as well. For instance, PCMM could support nontraditional services like VOD. Godas: Traditionally, if we look at our video experience, there has been a one-to-one relationship between the application or the video pump and the QAM. We learned from the lessons of the past that PCMM helps us break open that deadlock so that we can insert multiple applications on the QAM, which in this case is the DOCSIS bandwidth. Emmendorfer: The premise is to get convergence of the PCMM device to manage not only (the) DOCSIS layer but also the edge-QAM layer. Wouldn’t it be great as DOCSIS 3.0 and the next generation CMTSs come out if this PCMM platform can be supported for all of the services that would go across an infrastructure? CT: Obviously, PCMM is not just about voice. How can the technical community ensure that that’s not going to be the only focus? Emmendorfer: I would say service enabling. Once we enable services using PCMM, it will quickly emerge as a nonvoice play. CT: What services? Emmendorfer: Could be ensuring gaming quality of service. Could be for business services that we can offer using PCMM technology. And it could be video chat or videophone. Godas: Some of the transparent things that aren’t exactly, let’s say, billable would be mitigation of infrastructure impacting events—the whole plethora of things that happen to the user community. They’re getting attacked and (subsequently) taking down the infrastructure. We could use PCMM to slow this down a little bit, steer (users) to a place where they can get educated that what’s happening to them is hurting their neighbor. CT: Doesn’t that entail an intrusion detection system, which is somewhat distinct from PCMM? Emmendorfer: It could complement it. We’re deploying IP systems today for our revenue generation network to ward off network-borne viruses. It’s a level of bandwidth management, (beyond that used for peer-to-peer applications). It’s managing our bandwidth for things that are unknown or things that may happen, like network-borne worms, etc. What Joe’s referencing would be a way for us to enforce a policy to that customer—for instance, change their configuration settings to reduce their speeds, or even invoke a policy until that customer fixed their problem. Godas: Those systems that you’re talking about, a lot of us have them in place even at the Internet border to help us make peering decisions. And those very systems are there watching trends that show that someone is the target of a distributed denial of service attack. That could basically send a message to the policy manager and say, "Hey, I need to slow this person down or put them in a corner and let them know that they need to clean up their machine." Maloney: The intrusion detection type of applications that you’re talking about are actually very complementary to PCMM. We’ve been demonstrating for a number of months now applications that (use) the traditional intrusion detection application and then turn around and talk to the policy server to request quality of service so you can throttle it down over the HFC plant, as opposed to only doing it further back in the IP core. Delaney: Going back to your question about what kind of applications beyond voice PCMM be used for, there are a number of applications that operators have been talking to Camiant about—commercial voice, video telephony, video instant messaging, worm mitigation, games, turbo button, bandwidth on demand. Maloney: Streaming video. Delaney: Streaming media is an excellent one. There’s a lot of interest in using a PacketCable Multimedia policy server to enable high-bandwidth, high-quality video where you now have full screen, television quality over your high-speed data network. Maloney: I don’t really think it’s an issue of the technical community. It’s really going to be services driven. What’s going to bring in either the operational savings or the incremental revenue that’s going to drive it? PacketCable is a mature technology. It’s being deployed now, and I really think that, in the short term at least, that’s what people are going to focus their voice deployments on. And you’ll start to see that augmented with SIP-enabled PCMM, but I don’t think that’s going to replace anything in the short term. There’s too much effort, time and blood, sweat, and tears sunk into the PacketCable. CT: Let’s talk a little bit about SIP. Obviously PCMM has the ability to drive SIP. Should cable operators stop deploying PacketCable if they already are, or skip right to SIP? What’s your opinion? Godas: The PacketCable specification is going to continue being very important. I don’t think that one precludes the other here. I think that the ability to generate revenue is something that you can’t ignore. You do what you can with what’s available. You have embedded MTAs right now that are a great source of revenue that can be coming in, and there’s no reason why we should deny ourselves profiting from SIP-enabled end points. PacketCable Multimedia is the first chance we get to get out there and take hold of that. Delaney: I think there’s definitely an opportunity for both, and again they’re complementary. Many of the operators are looking at both SIP as well as NCS voice. CT: OK. We talked about the pros; let’s talk a little bit about the cons. What are the obstacles facing large-scale deployment of PCMM? Maloney: Plain and simple, it’s new. It’s going to take some time to mature. It’s going to take some time for it to be out there, stable enough, in production networks for people to be comfortable that it’ll scale. I think that’s probably the biggest short-term obstacle. From a long-term perspective, I don’t really see major obstacles to it because, at a fundamental level, it’s a very simple thing. All it does is it provides QoS to nonQoS-aware applications. It’s a very nice building block that you can build arbitrary applications on top of. Godas: We basically have everything we need to get going today, and we could have a significant deployment of it. (There’s obviously some work to be done.) We have this QAM bandwidth that moves around a little bit. Based on the modulation you use, you could have 10 megabits or 5 megabits. We’re going to have to continue working on how to reflect that bandwidth back to the policy managers so that it constantly knows what’s available. And if you have a business policy that says I only want 10 percent for gaming and 20 percent for XYZ application, well, 10 percent of 10 megabits is different than 10 percent of 5 megabits, and we want that dynamic relationship to constantly be established. We have the means by which to keep on top of that manually, but moving forward we’d want that to be automatic. And this is on the list of a couple of things that we continue to work through, but we can start now and do very well with it. Delaney: Virtually every operator has had a trial of PCMM in 2004, starting in January of 2004. 2005 is about deployments. Operators have been doing some planning about what applications they plan to deploy with PCMM. There’s a robust marketplace of policy servers available. Certainly Camiant has one available today. And then, of course, you need the CMTS code, and all the CMTS vendors have been supporting PCMM. Cisco’s been supporting it for a year now, and it has been in lab trials. It’s in a state of being ready to roll out. Of course, it requires a DOCSIS 1.1 network as well, which many operators have been preparing for. CT: We touched on this a bit, but just conclude: What’s next? Where does PCMM go from here? You talked about 2005, but let’s talk even beyond that. Delaney: Well, prior to the beginning of the panel, we were discussing dual-mode phones and the (integration) of wireless with broadband. I definitely think that PCMM has an important role there. When you’re in your home, you’re using the broadband pipe for your access to your telephony services, and then when you’re mobile, you’re using the wireless network. I think that PCMM is a very nice complementary technology to enable that dual-mode phone service to be rolled out. Maloney: From an application perspective, I really do think the sky’s the limit. As many different operators as there are out there, I think you’re going to see that many different applications. Every operator that I’m working with has something different that they’re focused on. I think Ed mentioned the dual-mode phones. I think that’s going to be one of the primary applications that’s really going to help drive both PCMM as well as the voice deployments for the operators. I actually think this is the thing that’s going to finally put videophones into people’s hands. We’ve been hearing for years and years how great it’ll be to have videophones. I think this is something that’s finally going to make that a viable alternative. My own personal theory is you’re going to start seeing videophones integrated into set-top boxes, making use of the TV and the sound that you already have. Simply adding a Webcam and a microphone to the set-top, you’re going to be able to really put (videophones) into the mass market. I think that you’ll see enhancements to operations that make this whole thing more manageable. I think that we talked about tying (together) the policy management of not just the QoS but also the edge QAMs. I think you’ll see a converged infrastructure for that. And last but not least, I think this is really what enables that triple or maybe even quadruple play where you really actually start to bring all these things together onto a common infrastructure. Godas: We’re really excited about the opportunity with policy management to be able to look at that bandwidth and schedule it with a larger degree of control across all the business units that share DOCSIS bandwidth. We’re going to be bonding those downstreams, etc., and having larger pipes to deal with. It’s going to become important to know which business unit is getting what portion of it to schedule properly. It’s kind of like the centralized booking agency. You don’t want to have the problem of one travel agency booking all the seats on the plane, as it were. (Regarding) the integration of the services, we’re looking to add value to our own content that’s already going to the TV sets and pushing it down to our data subscribers. Let them be able to watch TV on their PCs. Let them listen to the audio streams that we’re currently offering on our set-top box. We’ve already got initiatives this year to make that available. PCMM would only enhance that. These are all things we’re hoping to take advantage of (by the) tail-end of this year and into the future. Jonathan Tombes is editor and Laura Hamilton is consulting editor of Communications Technology. Did this article help you? Email comments to firstname.lastname@example.org.