With a target date of second half next year for initial trials and commercial rollouts in 2012, the company believes it has enough clarity on what MSOs want at this point to commit considerable resources to a next-generation device it calls the Video Edge Services Platform. The goal is not merely to facilitate lower costs for delivering more content in IP; it’s to provide a means by which services can be personalized with a high proportion of narrowcast components in a single chassis that does away with traditional service silos.
“Really, what we’re doing is looking at the personalization of content and being able to transport that across your mobile device, your PC and your television,” says Joe Cozzolino, senior vice president and general manager for access networks solutions at Motorola. “It involves everything from the software and hardware at the core all the way through to the device in the home.”
It’s a technically daunting business-driven agenda that aims to provide cable operators a means of adjusting service models in order to maintain their status as an essential resource to subscribers. “We look at this as a key enabler for operators not to become the dumb pipe,” Cozzolino says. “You have to have a real end-to-end solution around being able to do that so that you can deliver value to the consumer.”
As consumers access IP streams for ever more entertainment, personalization or a la carte selection of what they want when they want it wherever they are is becoming second nature, marking a stark contrast to the restrictions on choice and convenience imposed by traditional modes of service delivery. The challenge for cable operators is to transform their networks into personal service engines that do a better job of giving consumers what they want than they can find through alternative sources.
Comcast has become the most vocal MSO in the effort to meet this challenge, having touted its next-generation access architecture concepts worldwide. Now, as reported in the online journal Light Reading, the MSO is nearing completion of hardware specifications for what it calls the Converged Multiservice Access Platform (CMAP), with other, more detailed specifications pertaining to modular configurations and system management to follow.
Cozzolino credits Comcast’s thinking as a major spur to a recent acceleration in MSO timetables for transferring ever more content to IP. “They’re saying an entire architecture to enable Internet television has to occur,” he says. “That’s raised awareness dramatically over the last several months.”
While accommodating basic concepts of the CMAP in its VESP platform, including increased QAM (quadrature amplitude modulation) density and a shift away from traditional CMTS (cable modem termination system) architecture, Motorola is going further by creating a box that incorporates the configuration and bandwidth provisioning for all services, no matter what their transport formats might be, into a single unit, including business services that might be delivered over GPON or EPON fiber access links.
“It’s a fundamental architectural change from the core of the network all the way to the home,” Cozzolino says. While it’s important to lower costs and provide a flexible means of adding ever more IP streams, operators also need to be able to deal more efficiently with adding and changing services in the MPEG-2 domain, he notes.
“One of the biggest operational challenges operators have is the complexity of the headend,” he says. “So, whether it’s all the RF redundancy, the cross connecting, etc., what we want to see eliminated or significantly reduced is that huge coax nest in the headend.”
Today even adding an HD channel on the MPEG-2 transport side can be a time-consuming headache. “You need to do that more dynamically going forward, because there’s going to be so much content that’s going to be changing and coming dynamically, including over the top as well as linear and traditional on demand. What we’re architecting around our next-generation platform simplifies the headend dramatically.”
Even within the IP domain there are complexities that must be addressed at this point of intersection with the access network, Cozzolino adds. “Those packets have to be enabled whether it’s coming off of the wireless network, a DOCSIS network, a PON network or Ethernet,” he says. “You have to be able to process those packets and allow those packets to go across your mobile device, your PC and your television.”
Integrating PON-based services into the VESP control domain will facilitate using fiber to service businesses, Cozzolino notes. “Today operators can use the physical HFC plant to provide a low-cost means of supporting fiber access to businesses, but it’s still an entirely separate access network from an operational and management standpoint,” he notes. “You need to be able to provide that capability so that you don’t have disparate platforms.”
The VESP will collapse the functionalities of CMTS blades, QAMs and Ethernet switching, along with capacity for a terabyte of cache storage, into a unit that features hundreds of QAMs per blade, packet processing at 400 gigabits-per-second and 100 gbps throughput on the access slots. Today processing on a per-slot basis runs in the range of two to four gbps.
Motorola’s move away from the traditional CMTS in conjunction with Comcast’s call for similar DOCSIS architecture in its CMAP specs appears to settle the question of whether a centralized CMTS structure would address the needs of high-volume IP video throughput over bonded DOCSIS 3.0 channels. “From my perspective with a business that owns the CMTS, we do believe we have a dramatic shift away from the existing architectures,” Cozzolino says. “I don’t even like to call it a CMTS anymore.”
While some suppliers assert ever-increasing processing densities of microprocessors will allow existing CMTS blades to accommodate high volumes of downstream video cost effectively, Motorola has concluded the proportions of video will be too great to justify that approach, especially in the context of the efficiencies to be achieved through marrying the CMTS functionalities with Ethernet routing and QAM processing in one box. Today, Cozzolino says, about 90 to 95 percent of the cable IP traffic is transported as DOCSIS data and the rest going for packet voice. In the future, the volumes of traditional data and voice will remain more or less the same, but together they’ll represent less than 20 percent of the cable IP traffic, with the rest going to video.
“You’re looking literally at a logarithmic change in terms of capacity, density and throughput in this next-generation architecture,” he says.
A consolidated edge platform like VESP will give operators flexibility to continually shift content from MPEG-2 to IP and MPEG-4 and to allocate more bandwidth for dedicated unicast service streams as network support for time-shifted services intensifies. Such flexibility goes hand in hand with having enough per-user bandwidth on the coax end to support IP unicast and multicast services.
Motorola anticipates operators will be able to do this by leveraging its and other vendors’ fiber deep technology as a cost-effective way of lowering the number of homes served per node.
While Motorola is a leading supplier of PON systems, including products for the Verizon FiOS network, Cozzolino believes the IP transition can be accommodated without resort to going all the way to the home with fiber.
“I don’t believe you’ll see fiber right to the home anytime in the next five, six, seven years,” he says. “But I really do think fiber, from a business services perspective, will be critical to the operators.”
Motorola is undertaking VESP development as a “black-box” project apart from its existing CMTS engineering efforts. But the company’s existing Integrated-CMTS platform will benefit from the new initiative, allowing operators to leverage DOCSIS-related advances within their existing I-CMTS infrastructures.