The far-reaching step beyond switched digital video (SDV) is rooted in the company’s field-proven ability to coordinate complex operations on multiple network components in real time, says John Holobinko, vice president of marketing at BigBand. “We are executing 110 million real-time transactions per day across 24 million households using this model,” Holobinko says. “What we’re announcing now is the next logical move beyond SDV with our control plane technology.”
BigBand is bringing its Converged Video Exchange (CVEx) edge platform to market at a moment when the complexities of moving to the next level of services, applications and advertising appear to be overwhelming the old siloed approaches to service management and the network’s capacity to deliver HD and SD content over RF and IP streams in dynamically changing combinations of broadcast, multicast and unicast modes. The CVEx is designed to overcome these challenges, but it will require changes in approaches to some things such as how IP video is delivered and how bandwidth is allocated which may be hard for some engineers to accept.
Nonetheless BigBand is confident reductions in complexities that can snarl operations and hold back new service launches combined with significant bandwidth efficiency gains will persuade operators to take a new architectural approach that is minimally disruptive to what they’re already doing. “Transitioning to next-generation networks capable of seamless video delivery to multiple networks opens up tremendous complexity,” says BigBand COO David Heard. “Our job is to arm our customers with a more simplified and nimble video delivery solution that consistently delivers a quality video experience to multiple devices, anywhere.”
Slated for first-phase rollout in the fall, the CVEx removes the boundaries between disparate video service types by dynamically controlling the allocation of all RF and IP video services within a single bandwidth pool, Holobinko explains. As a result, he says, the control plane, running on fully redundant rack-unit modules or blade servers, eliminates the dedicated bandwidth silos for individual services such as VOD, SDV and IPTV while simultaneously improving spectrum utilization. Moreover, he adds, the platform provides subscriber and video service measurement tools that cross multiple data planes with holistic views of program and service behaviors, network resource usage and “what if” scenario planning.
The idea of using network edge intelligence to dynamically allocate RF bandwidth on universal edge QAMs (quadrature amplitude modulators) has been in play for some time in the cable industry, but with limited implementations. With the introduction of DOCSIS 3.0 specifications for the modular CMTS (cable modem termination system) the industry has been given an option to use those universal QAM resources to carry IP traffic in combination with the TV streams.
But all of this convergence through the QAM assumes the individual services remain separately managed within their silos with no coordination at the session management level. “What we’re talking about with the CVEx is different from the role played by the ERM (edge resource manager), because the ERM only deals in RF,” Holobinko says. “And it can’t combine sessions.”
A fully converged video environment requires a unified control plane that can coordinate all the network elements to achieve high transaction rates with fast response times on multicast and unicast session requests, Heard says. “By unifying the video through our intelligent software control plane, CVEx can help transform today’s architecture into a true converged video solution delivered to any device so that in the future, it won’t matter if it is wired or wireless, legacy or next-generation IP,” he notes.
The CVEx control plane determines the best way to satisfy requests from users as they’re watching services. “The control plane allows resources to be allocated as needed in response to user requests,” Holobinko says. “It’s instantaneous, like SDV, but it extends management beyond RF into the IP domain and applies to unicast as well as broadcast and multicast delivery modes.”
For example, Holobinko notes, someone may be watching a linear broadcast or multicast program on a service that allows users to pause linear shows and pick up where they left off when they return to the TV set. “When you hit pause you’ve made the jump from linear to unicast mode, which is seamlessly and instantaneously handled through the control plane,” he says.
This high-speed intelligence has important implications for advertising as well, Holobinko adds. If an operator wants to implement a personally targeted addressable advertising capability that requires unicast delivery of different ads in a given time slot, it won’t be necessary to consume the bandwidth it would take to switch the entire linear program to unicast. Instead, the control plane can grab available bandwidth wherever it might exist across the 860 MHz spectrum to accommodate unicast delivery of just the 30-second addressable spots to all the tuned-in users and then instantaneously switch the program back to broadcast or multicast when the ad ends.
Such capabilities illustrate the depth of awareness into what is going on at the set-top and across all network elements that is required for this architecture to work. This access to and accumulation of data for on-the-fly processing is a key benefit that applies to measuring ad performance, Holobinko notes. Because the control plane operates seamlessly over the legacy and IP video domains, operators can gain an aggregate view of advertising performance no matter what devices content is viewed on, which will be instrumental to delivering ROI on new advertising service models.
“The new wave in television is video is video,” Holobinko notes. “I don’t care what the headers are on the digital content. I want a single means of delivering everything to different devices.
“When the device requests service, that request goes to the session manager, which goes to the CVEx and asks what that device needs,” he continues. “The CVEx knows what the device attributes are and so links the session manager to the appropriate resource for meeting those requirements and makes sure the mode of delivery and appropriate bandwidth requirements are satisfied as well.” Similarly the CVEx makes sure the ad servers and back-office elements are tied into the session manager for fulfillment on the business requirements of a given content stream.
Initially, with the fall release, the CVEx will support shared bandwidth operations among RF streams, leaving the integration of IP streams into the uniform bandwidth management process for implementation in 2010. To facilitate operator migration to the new platform, the solution will be offered modularly so that operators who just want an ERM and a single application capability such as SDV can obtain the platform at cost parity with legacy devices. Rather than being limited by the capabilities of the typical ERM the operator will have a core platform into which he can plug the full range of functionalities the CVEx is designed to support, Holobinko notes.
The major bandwidth-savings capabilities of the new platform will come into play with integration of the IP flow into the unified spectrum management system. Instead of implementing bonded channel DOCSIS 3.0 allocations over dedicated DOCSIS channels, the CVEx bandwidth management system will bond as many RF channels as the operator wants to simultaneously transmit IP content in the spaces left open as the TV content bit rates fluctuate.
“Legacy set-tops look at the RF TV signals, but ignore the wasted bandwidth in those channels,” Holobinko says. “With the control plane I can define any of those channels as part of a bonded channel group for my IP traffic. That traffic doesn’t care about what else is there. The stat muxing system will allocate whatever bandwidth isn’t consumed by the RF TV to the IP video.”
Engineers assert that with statistical multiplexing and variable bit rate recording the bonding of channels for delivery of IP video can cut bandwidth consumption over those bonded channels by up to 25 percent in comparison to the bandwidth that would be consumed by streams dedicated to specific 6 MHz channels. But, Holobinko notes, if an operator is only bonding 16 out of the 128 available 6 MHz channels, that savings on the bonded channels translates to just a 3 percent savings across the entire spectrum.
By spreading the IP traffic across the RF TV channels, he says, the bandwidth savings is greatly expanded. The upshot, he adds, is that operators will be able to support a jump in the unicast channel count as they introduce more time-shifted services without having to reduce the 500-home node serving areas that characterize most large cable systems.
Holobinko notes other benefits of the unified control plane include the ability to operate user navigation guides from the network rather than lodging them with each individual device. This will allow operators to maintain consistency in design, content options and functionality across the guides for all devices as the video service expands beyond the set-top.
BigBand is also looking at implementing IMS (IP Multimedia Subsystem) proxy capabilities in the control plane, which would take some of the intensive processing burden off consumer devices as the industry applies IMS functionalities to converge mobile and fixed services. The CVEx strategy will also have implications for Tier 2 and 3 cable operators in the year ahead, when bandwidth requirements for broadband and the need for SDV support will be key issues. “For operators who don’t want to spend on the costs of upgrading 550 MHz plant but also can’t afford the clamping and rate shaping costs of SDV, the CVEx with its bandwidth management efficiencies offers a better option,” Holobinko says.
Moreover, he adds, many of these smaller operators have older CMTSs which don’t lend themselves to DOCSiS 3.0 upgrades. Rather than replacing the CMTSs to achieve decoupled downstream/upstream capabilities essential to cost-effectively supporting higher volumes of IP video streaming they will be able to make use of the vIP Pass technology BigBand introduced earlier this year (see March issue, p. 8), which is an intrinsic component of the CVEx support for integration of IP video into the universal control scheme. This solution encapsulates IP video into DOCSIS frames at the network edge, eliminating the need to pass those streams through the CMTS.
The vIP Pass approach to utilizing DOCSIS 3.0 modems and functionalities represents one of the conceptual hurdles engineers will have to clear if they are to make full use of the spectrum efficiencies offered through the CVEx. Bonding was not part of the initial vIP Pass release but presumably will be in conjunction with the incorporation of the IP stream into the CVEx universal bandwidth management scheme next year, which, as noted, depends on channel bonding capabilities.
As previously reported (April, p. 1), the industry through CableLabs has begun to explore a new approach to DOCSIS that would work better for streamed video than even the DOCSIS 3.0 model, which was designed to leverage the existing CMTS architecture. Holobinko says discussions revolving around an IP termination platform as the successor architecture would be a perfect fit with the CVEx architecture, serving to eliminate the vIP Pass approach to solving DOCSIS 3.0-related conundrums.
BigBand plans to demonstrate the CVEx at the CableLabs Summer Conference in early August. Holobinko says the demo will lay to rest competitors’ contentions that the scheme is not practical. BigBand sees existing SDV customers as a strong starting point for selling the technology.