Telecom network optimization for video services faces an uncertain future as carriers find that the explosion of both fixed and mobile video use is not generating commensurate growth in video revenue. With no corresponding revenue growth, few carriers are willing to spend money to optimize their networks for video.
Cisco released an (admittedly self-serving) report projecting that Internet traffic will quintuple by 2013, driven largely by various types of video, which Cisco estimates will make up 60% of all traffic.
The problem is even more acute for wireless service providers. Since many cell sites are served by costly leased lines, video downloads are exponentially more expensive to pay for over the network.
Each cell site's limited wireless throughput means that a handful of video users accessing the same tower can easily degrade the speed and quality of all other users utilizing that tower, resulting in angry customers and higher churn.
Even though network operators have several tools available to increase the reliability and optimize their networks for transmitting online video, such as video compression and point-to-multipoint MPLS, there is little incentive to invest.
Ultimately, service providers on both landline and mobile fronts are facing the same fundamental problem: Video services are an incredibly tricky proposition to make money from, even in the best of times. With the economic recession, a vague business strategy of "ad revenue" just does not pass muster, even as consumers continue to expect the content free-for-all they have had the past several years.
Mobile video services: Consumer demand, but business case comes first
Consumer and business demand for such video growth is quite understandable: Hulu, Skype and YouTube have all been popular among home users, while video conferencing has come back into favor with companies looking to cut travel budgets.
While techniques exist to help optimize this traffic and cut costs, analysts and former video optimization executives alike offer the same advice: The business case must come first.
"When you look at how this is going to work, and you assume the service provider is going to provide smartphones for the customer, and they watch a Hulu video, who is going to pay to make that work?" said Tom Nolle, president of CIMI Corp. "The answer is nobody, because the service provider has no financial incentive to do it; Hulu isn't going to do it because they don't have the money, and they don't control enough of the path to guarantee the experience even if they did."
Nolle said that leaves one very fundamental question for video optimization solutions: Who is going to pay for it?
Even video compression solution is a tough-sell
Solutions exist to cut video transmission costs. One of the simplest, most direct methods is more efficient video compression: With less data to send over the network, bandwidth costs for service providers are lowered, and the chance that streaming video will degrade network performance also lessens.
TruVideo's video compression technology promised just this, but the company has had a hard time actually closing deals with what seemed like a win-win proposition for both service providers and content providers that TruVideo helped to broker deals with.
While still in business with a few active customers, including independent mobile movie service mFlix, available through Sprint, TruVideo's business is now winding down, according to founder and chief technology officer Avideh Zakhor.
Existing deals are still being honored, but Zakhor is not optimistic about new business coming about.
"It became a licensing game rather than a high-profit, high-margin sort of business proposition," she said. "We tried different business models and really the dollars could not be justified."
Things looked quite different in 2000, when the company launched.
The prospects of helping the large carriers work out deals to get in on content providing seemed like an untapped opportunity, and Zakhor said TruVideo managed to engage in high-level talks with Verizon and major content providers.
All of the deals TruVideo eventually made involved the end user paying a monthly subscription, Zakhor said, but after each participant in the deal – the content providers, the service provider, the lawyers for each party, etc. -- had a cut, there was little revenue to go around.
"It was a difficult business case to make even though the technology was good," she said. "It's not an easy business to be in."
She recalled hearing the grim joke in her industry that wireless video is an oxymoron: Someone, whether it's users, the service provider, or the content provider, ultimately pays a high premium to deliver video that, constrained by wireless bandwidth, is of lower quality.
"And the biggest viewers of video are kids and teenagers who don't have a lot of extra cash," she said. "Coming out with business models that make sense is crucial. We tried a lot of different models but none of them stuck because of the constraints in terms of the service providers' profits, and the content providers' profit."
MPLS video optimization faces hurdles
Similar hurdles exist for the adoption of point-to-multipoint MPLS implementations, an optimization technique that would at least be driven by a service provider profit center, IP video. MPLS technology could cut bandwidth consumption within a service provider's network, as one content stream could then be propagated to multiple users as needed, rather than having a service provider send out unique streams to each user.
But the technology's real value, Nolle said, is in improving the quality of service (QoS) controls that service providers have on IP video content streams.
"Video distribution is the only credible multicast application with potential revenues associated with it," Nolle said. Service providers who broadcast IPTV can benefit because, by its very nature, broadcast is multi-point.
But even one of the technology's biggest backers, Juniper Networks, admitted uptake is slow.
"I will say that a fair number of carriers currently don't use point-to-multi-point MPLS, and the reasons for that vary," said Scott Shoaf, a member of Juniper's strategy and planning group. "A lot of times it is the religion or beliefs or technical vantage point around MPLS as a whole to begin with. There are networks out there that don't run MPLS at all, as an example."
But Nolle said the reluctance to fully embrace the potential network optimizations that point-to-multi-point offers throughout service providers' networks extends further than religion to business realities.
For service providers, it's a delicate balance of the cost of infrastructure compared with the expected rate of return.
In Verizon's case, for example, the demand density is high enough to support a fiber-to-the-home (FTTH) FIOS network, which is conducive to analog rather than IP transmission of video.
Smaller service providers, on the other hand, do not have the economies of scale to make running MPLS throughout the network profitable.
"If your demand density is too high, it doesn't work," Nolle said.
But he said that there are still some MPLS opportunities in telecom network optimization that almost all service providers could take advantage of.
Nolle said that as more service providers invest in content data networks (CDN) to act as way-points for caching video and other large content, point-to-multi-point could ensure that all content is streamed to these CDNs in a timely, accurate manner, since service providers want data arriving at the same time at CDNs throughout the network.
"The CDN application is going to become more viable over time," he said. "The CDN application works for everybody, but the larger your geographic scale, the better it is."