Mobile TV and video services, for all the hype and interest they are generating among TV networks and movie studios that want to exploit new ways to deliver their content, remain applications whose technology foundations still are shifting into place under them.

Most mobile TV/video services today are based on a unicast model — applications downloaded from the Internet. But, within the next few years, networks based on DVB-H, MediaFLO and other technologies will be sprouting worldwide, many of them built by third-party companies, rather than the traditional mobile carriers. These networks will bring mobile TV/video into the broadcast age.

In total, mobile TV/video will represent a $6.5 billion market opportunity by 2009, according to Jeff Heynen, broadband and IPTV directing analyst for Infonetics. How much of that will be derived from unicast services and how much will be from broadcast services remains to be seen, but it's safe to say that regardless of the network architecture used to deliver mobile TV/video, carriers will have to meet a quality-of-service benchmark that has been heavily influenced by traditional home-based TV viewing.

“Mobile video needs to replicate the home TV experience as much as possible, be it over broadcast (radio frequency) or unicast (3G) mobile devices,” Heynen said, in a statement announcing his latest research. “The mobile video services available today leave a lot to be desired, but people are still subscribing because they're excited about the future of the technology. The convenience it provides outweighs its limitations.”

But vendors involved in building and testing video networks and services say that beyond the basic understanding of what customers may be looking for, a mobile video quality benchmark has not been firmly established.

“There's no detailed experience or research yet available to know what people are expecting of video from a quality standpoint,” said Dave Glidden, director of TV strategy and business development for Harris Corp. “The nearest assumption is that it might be something similar to TV quality in the home, with a reasonably consistent frame rate, as little blockiness as possible — we'll all learn eventually what the minimum accepted levels are.”

Meanwhile, what is known for certain is that mobile video also poses new challenges for how to test for and manage quality. Regardless of whether mobile video is being transmitted in a unicast or broadcast architecture, “The key thing with video is that has to meet requirements — it has to be reliably transmitted at a constant bit rate, and it has to be transmitted in a ‘loss-less’ environment,” said Paul Robinson, senior manager of MPEG for test vendor Tektronix.

In the course of launching a mobile video service in a unicast environment such as a UMTS network, carriers might rely first on out-of-service tests, which would use simulated network elements carrying video traffic. Network infrastructure would be replaced in these tests by test set elements that would inject video traffic into the radio network controller and other elements, according to Othmar Kyas, director of strategic marketing and network diagnostics for Tektronix. “This is part of the process for preparing network infrastructure for the real-world mobile video environment,” he said. “There are also in-service tests where you do passive monitoring of the control plane and the user plane to compile an in-depth quality analysis.”

In a broadcast network environment, such as a DVB-H network, tests would focus on analyzing the output from the RF transmitters that would broadcast the video to multiple users. Since broadcast networks have a multi-casting architecture dedicated to video, there is less concern about network loads and peak usage patterns and more focus on signal strength and quality.

Glidden, of Harris Corp., noted that with a push DVB-H service, it will be much easier to control video quality. If computer models and real-world experience determine that a signal reception is lacking in a certain geographic area, transmitters can be moved or re-configured to cover the gap. “The broadcast network doesn't worry about network loads or whether you have one subscriber or a million of them,” Glidden said.

But the ultimate measure of quality, whether in a unicast or a broadcast environment, will be defined by what the earliest generations of mobile video consumers think about their services. Managing the subscriber experience picks up on where the traditional network and client test sets leave off.

Test vendor Spirent last year acquired SwissQual, a small company that had been doing pioneering work in the area of subscriber experience management algorithms. “The core of subscriber experience management are complex speech and video measurement algorithms,” said Hesham ElHamahmy, Spirent's senior director of business development. “Basically, it was developed by putting groups of people together in a room to see what they think about various aspects of video quality and then creating a mathematical model to measure that.”

The algorithms are designed to measure not only video quality, but also perceived speech-to-video synchronization. SwissQual created its user profile group by using guidelines from the ITU. It assembled groups of at least 24 people of different backgrounds and usage experiences to view video clips that had been captured live and that the group participants had never before seen.

Pero Juric, chief technology officer for wireless at Spirent, who arrived via the SwissQual acquisition, added, “If you can build different kinds of databases — a database for each type of degradation of video quality, then you have created a new, objective model for how to test mobile video quality. You can create new test solutions that combine existing video quality tests with the new algorithms that are based on what the human eye and system perceive.”

Juric said Spirent is now working with the ITU to help establish what the company called a “conversational multimedia standard.”

1 MILLION

Estimated number of mobile video customers worldwide today¹

250 MILLION

Number of mobile video customers predicted worldwide by 2010¹

$27 BILLION

Total expected worth of worldwide mobile video market by 2010¹

22 MILLION

Number of mobile video customers expected in the U.S. by 2010²

13.2% MILLION

Percentage of U.S. customers indicating interest in subscribing to mobile video services²

Sources: ¹ ABI Research, ² In-Stat