NGNA

Most network engineers would agree that the modern cable plant is a marvel of efficiency. Through innovative design and dramatic leaps in technology over the past few decades, operators have evolved from basic analog video delivery services to state-of-the-art communications enterprises capable of delivering advanced digital voice, video and data services. But even though the technical advances have been dramatic, most cable operators realize that they must compromise on efficiency in order to cram three separate services onto a single network. This silo approach-one silo each for voice, video and data-essentially forces operators to build what amounts to three independent networks in order to shuttle bits and bytes between the home and the outside world. Wouldn’t it be great, then, to have a single overarching Internet protocol (IP) network architecture that would allow operators to deliver all types of communication services at once? Enter what is being called Next Generation Network Architecture (NGNA). What it is "Next generation architecture basically converges all signals onto an IP infrastructure," says Tej Kohli, director of service provider marketing at Cisco. "We will no longer have to build a siloed architecture for each service. The commonalities (of the data)-first at the core, then the edge, then the common access-allow for much greater (bit rate) reduction and greater flexibility in service rollout." Although the idea of a unified network scheme has been kicked around for several years, the NGNA ball got rolling in earnest in 2003 with the creation of a limited partnership, the NGNA LCC Group, consisting primarily of Cox, Time Warner and Comcast. Technical staff from the three firms issued a request for information (RFI) to various vendors more than a year ago. This past October, the group formally handed off further development to CableLabs. Details of the forthcoming system are a closely guarded secret, but the goal is to develop a series of new edge devices that integrates a high-density quadrature amplitude modulation (QAM) device, a Data Over Cable Service Interface Specification (DOCSIS) cable modem termination system (CMTS) and video processors. When incoming data from the Gigabit Ethernet (GigE) backbone hits the edge, the NGNA platform identifies the type of signal it is and routes it to the appropriate device. Ideally, NGNA will drive down equipment costs even further, free up valuable bandwidth and usher in a wealth of new niche services to boot. The mere prospect of IP video services long has been a pipe dream for the industry, but with NGNA, it finally may offer a real chance for operators to shelve the leased cable box business model for one in which consumers purchase their own devices at retail stores. One need only look at the raft of low-cost consumer IP devices that take advantage of telco IP networks (devices that may soon include video decoders as well) to imagine all the possibilities. "(NGNA) will build a series of bridges between the home, the cable plant and the wider network to allow multiple devices to work together seamlessly while still providing the same quality of service," says Ralph Brown, CTO at CableLabs. But don’t expect CableLabs to initiate a formal NGNA working group or even to deliver a final NGNA spec. Rather, Brown says, the organization will incorporate the NGNA concept into new specifications, as well as future iterations of existing specs, such as PacketCable and DOCSIS. "There will be no NGNA spec per se," he says. "There will be a refined direction to our existing projects. We will define optional interfaces, but we don’t want to mandate a specific implementation. We don’t want to close off innovation and potentially create more costly solutions." Costs Reducing costs is always a key driver in any new development, and NGNA is no exception. However, there is some concern that while the technology might cut costs in some areas, it could increase costs elsewhere. "There will be substantial savings in hardware associated with the transport of information," says Stan Brovont, vice president of marketing at Arris. "There will be operational savings in monitoring and maintaining networks." "But the market for services that this architecture will support already is very mature," he added. "Most operators have invested huge amounts in edge products to support voice, video and data. By the time these elements are redesigned for the new architecture, it’s not clear how much demand there will be. "Also, while we are making the core network simpler, we are making the edge network more complex. Today, there are maybe four or five major boxes on the edge. With NGNA, there could be 11 or 12. With more complexity, there are questions of interoperability, reliability, operations …." Still, Brovont says Arris is "enthusiastic" about NGNA and already has announced a device, the Q5 digital multimedia termination system, that he says provides a "naturally evolutionary step to NGNA." Some vendors are making significant strategic moves now in anticipation of a unified IP architecture. Michael Adams, vice president of technology and architecture at Terayon, says the company’s recent decision to exit the CMTS market will allow the company to better focus on network level convergence augured by NGNA. He says that an examination of bit transport economics showed that MPEG-2 services cost operators about $30 per megabyte, while DOCSIS, because it is a two-way protocol, runs upwards of $300 per megabyte. By focusing on a unified IP platform for all data, he is banking on a more cost-effective solution. "We didn’t make this decision lightly," he says. "We’re saying a network form of convergence is more interesting than a physical shelf level convergence. What’s driving NGNA is the fact that operators are looking at the economics of VOD and saying: ‘Wow, $30 per meg. I wish I could apply some of those economics to data.’" What the future holds However, other vendors appear to be hedging their bets because there is no clear consensus as to how NGNA might alter the networks of the future. In fact, there could be a radical shift in what the technology offers the cable industry in the near term and then in the long term. "The end point is difficult to define because NGNA is more of a journey," says Cisco’s Kohli. "It may be that there is no specific end point." Kohli argues that a key determinant in how NGNA will improve networks is the creation of a new service control layer, which interconnects the basic network layer and the applications being provided, and to what degree it will be standardized. "There needs to be a new layer of service control management to take the complexity away (from a unified network)," he says. "Service control answers clear questions, like: Who is the user? What authorization does he have? How does the user interact with the application? How are the home phone and the work phone linked?" In other words, a unified IP network will only be as good as the operator’s ability to control it. All of this will weigh heavily on the members of CableLabs where, if anything, there is a resistance to imposing a rigid framework. "In any technical development initiative, there will always be issues," Brown says. "On the DOCSIS, CMTS and cable modem side, we have the first round of proposals in from the vendors. We’re kicking off a team for spec development in December or January. PacketCable is undergoing a similar process, and OpenCable has efforts underway." "NGNA is so overarching that we have set up a significant amount of coordination on the business level, the CTO level and the technical level. Our hope is that these teams will be able to ride herd and make sure these overarching issues are being addressed." With some of the cable industry’s heavy hitters firmly on board, it seems unlikely that NGNA won’t become a reality fairly quickly. But exactly what form it will take and how it will ultimately affect the industry is still very much up in the air. Art Cole is a contributing editor for Communications Technology. Reach him at [email protected]. Security Is Key for NGNA One of the most crucial aspects of the Next Generation Network Architecture (NGNA) is content security and conditional access (CA). Without a rock-solid means to ensure content won’t be pirated, it’s very unlikely that cable operators and other network providers will get Hollywood on board. Not only that, a robust security system also will better allow cable operators to control and manage the content coming over their networks, maintaining their status as service operators rather than merely access providers. "Security would appear on the surface to be a small piece of the overall network architecture," says Brian Baker, CEO of WideVine, a Seattle-based developer of Internet protocol (IP) security systems. "But it’s by far the most important piece." Industry sources say that security is the one area that CableLabs is not overseeing. Instead, Comcast is said to be working on a system in-house, which it will present to the industry, pending further development. "Because of its critical nature, (security) is an area that has an even higher level of interest than other projects," says Ralph Brown, CTO at CableLabs. "Their (Comcast’s) leadership role is more significant than in other projects, but it isn’t something that is isolated. All of our members are participating." There is wide consensus, however, that NGNA security should do away with embedded hardware solutions and instead adopt a more dynamic version that can be downloaded from the network, perhaps using a smart card. Not only would such a system cost less than traditional hardware solutions, but it also could be continuously updated with new algorithms and security keys to thwart would-be hackers.