Tight squeeze for mobile TV

According to a recent report from Analysys, an independent research group in the United Kingdom, capacity on a third-generation, or 3G, wireless network could be exceeded as early as 2007 if 40 percent of subscribers view even eight minutes of video per day.

"Streaming video consumes 10 times the bandwidth over a network that voice traffic consumes," said Alastair Brydon, one of the authors of the report. "So watching 10 minutes worth of video per day will have a significant impact on the network. Right now, the 3G networks are empty, so it's not a problem. But if the service proves popular, then it could be a big problem."

Mobile operators have spent billions of dollars on 3G wireless networks in order to deliver new services such as e-mail, music downloads and video. In the United States, major carriers are just now in the final stages of deploying these networks. And as completion nears, carriers hope the new capabilities will help boost sagging revenues.

Video and TV services for mobile phones are being hyped as major moneymakers for the future. Three of the major U.S. cell phone operators have begun offering a mobile TV service. Verizon Wireless has its V Cast service. And Sprint and Cingular offer a service from start-up MobiTV. At the CTIA tradeshow in San Francisco last week, content providers such as MTV and Warner Music Group were busy announcing deals to deliver content via cell phones.

The stage has been set for a major market to emerge. But so far, mobile TV services have not been a big hit with consumers. Still, wireless carriers hope that the faster 3G network will improve image quality and lead to more content--such as movies, news clips, real-time sports, mini-soap operas and full-length TV programs--thus driving demand.

"We're very encouraged by the market," said Dale Knoop, general manager of multimedia for Sprint. "We definitely think there's a lot of demand to have things that entertain and inform you on the go."

But the truth is that mobile operators could face big problems if their TV services become popular too quickly, because 3G was never built to deliver streaming video.

"The carriers are going to have to move the traffic off the cell network eventually," said Albert Lin, an analyst at American Technology Research. "It just doesn't have the kind of capacity that video demands."

This is exactly what happened in South Korea when carriers initially rolled out streaming video services there. Within eight to nine months, the network became congested with video traffic. SK Telecom quickly realized that a new approach was necessary. So it built a separate satellite network to broadcast its mobile TV service.

Why can't 3G networks support high volumes of video traffic? 3G wireless networks are divided into cells. Users in a given cell share the available bandwidth. The networks are also designed to be "unicast," which means signals are transmitted between a single sender and a single receiver. If 500 people in the same cell decide to watch the same video clip, the network has to transmit a copy of that video clip over the network to each user.

This isn't a big deal when users are chatting on the phone, sending text messages or downloading ring tones, because those applications use up relatively low amounts of bandwidth. But video eats up roughly 10 times more bandwidth.

"3G is a poor solution for a big media event, like a breaking news story or a championship sporting event," Brydon said. "Just when everyone tunes in to see what's happening, the network fails. Not having enough capacity is clearly a bit of a weakness."

A more efficient way of delivering mobile TV would be to broadcast the content to users, and allow those who want to view it to tap into the network, he said. This approach, used in traditional broadcast television and radio, means that video clips are transmitted only once over the network, instead of being replicated and transmitted hundreds or even thousands of times.

Several technology companies and standards bodies are already working on solutions. One group, called the 3GPP/3GPP2, is working on modifying 3G. The technology, called MBMS (Multimedia Broadcast and Multicast Standard), will likely be available for W-CDMA and CDMA-2000 networks in 2007. The benefit of MBMS is that it doesn't require additional spectrum or licensing. What's more, coverage will be identical to conventional 3G networks.

But the drawback of MBMS is that it requires operators to set aside capacity that could otherwise be used to sell lucrative point-to-point voice or data services. Also, because of the capacity limitations, services would likely offer only a limited number of channels.

Other approaches call for building dedicated broadcast networks using technologies that would greatly increase the number of channels

The verdict is still out on which method operators and handset makers will adopt. DVB-H (Digital Video Broadcasting ? Handhelds) is already being tested in Berlin; Helsinki, Finland; Oxford, England; and Pittsburgh. Handset maker Nokia and Crown Castle, which sells wholesale wireless capacity to other carriers, are testing the technology in Pittsburgh. Crown expects to start offering the portable TV broadcasting service next year.

DMB, or Digital Mobile Broadcast, is a standard developed in South Korea. It works much like digital radio in that country.

And then there is MediaFlo, a technology developed by wireless handset and chipmaker Qualcomm. MediaFlo consists of an end-to-end network that uses new wireless transmitters and receivers. Qualcomm doesn't plan to sell video services directly to consumers, but will offer wholesale access to its network for providers offering mobile TV service. Mobile operators will be able to offer from 15 to 20 channels of broadcast-quality TV. The service is expected to be commercially available by the end of 2006.

The benefits of using dedicated broadcast technology are obvious, but it will require operators to upgrade portions of their network. It will also require customers to buy new handsets.

Another potential consequence of building these separate broadcast networks is that it could take some control away from the cellular phone providers. Because these networks could be accessed by anyone, content providers could sell their brand directly to consumers, cutting out the mobile operators entirely.

So far, mobile operators aren't saying much about their plans for the future, but many have acknowledged they are looking into new technologies.

"We've said publicly that we're looking at MediaFlo, but that's all I can say about it," Sprint's Knoop said.

But most experts acknowledge that it's unlikely mobile operators will face a problem in the short term, because mobile TV isn't likely to take off for at least another two years.

"MobiTV says they have 500,000 subscribers," analyst Lin said. "But that's out of a total of 175 million cellular subscribers. That doesn't sound like a significant market to me yet. Despite all the promotions and hype, I don't think the volume is really going to be significant until at least 2007. And by then, these new technologies should be available to the market."