Qualcomm Inc (QCOM)

[khemara_qc]

Vendors Investing Heavily In Mobile TV

By Rick Merritt Courtesy of EE Times

September 12, 2005
Page 1 of 3


http://www.mobilepipeline.com/showArticle.jhtml?articleId=170702174&pgno=1

You want TV with that?" The cellular industry hopes it won't be long before the average customer answers in the affirmative, making mobile television the first mainstream application for wireless data. Construction of more than $1 billion in chip sets, handsets and mobile broadcast networks is under way in anticipation of a market that could be in the tens of billions of dollars.


Market watcher Informa Telecoms & Media (London) predicts there will be more than 70 million TV-ready handsets sold annually and 124.8 million mobile-TV users by 2010. The Shosteck Group (Silver Spring, Md.), in its most pessimistic scenario, projects mobileTV revenues of $9.7 billion by 2010; its optimistic projection is $27 billion.


That said, a few realities keep the wireless mass markets of TV and cellular from spawning the hoped-for bonanza. Low-power, integrated chip sets are not ready. Handsets — even before they take on TV — are saddled with small screens, only fair battery life and mediocre reception.


Multiple network architectures may confuse the market as they compete to carve out a slice of the pie. Most of them have yet to complete successful technology and commercial trials; some lack adequate spectrum. And broadcasters still have little content tailored for mobile users.


"Carriers are in the due-diligence phase of evaluating the technologies," said Jeff Lorbeck, senior vice president and general manager for MediaFLO, a subsidiary of Qualcomm Inc. that will spend $800 million over the next five years to build a mobile multicasting network based on a new spec that it has yet to be released publicly.


"I think the carrier commitments will come very soon — probably before the end of the year in the U.S. — and commercial services in 2006 are a very likely reality," said Bob Shallow, a director of music and rich media for Nokia Corp.


More than 10 trials — in places ranging from Helsinki to Barcelona and Pittsburgh to Australia — are in the works for the DVB-Handheld multicasting network architecture that Nokia favors and that Europe's Digital Video Broadcasting Project (Geneva) codified as a standard in late 2004. "All the pilots have the chance of becoming commercial services," Shallow said.


"It's still early days, but the indications are that TV is the single wireless application with the biggest impact," said Phillip Alvelda, chief executive of startup Idetic Inc. (Berkeley, Calif.), which has operated a mobile-TV service over existing cellular networks since November 2003. The $10/month MobiTV service now has 450,000 subscribers.


"This is so good for carriers" that they're requesting new phones come preloaded with the application, said Alvelda. Idetic's service is offered by several carriers, including Cingular and Sprint in the United States and Orange in Europe, and Alvelda claims he has 30 carrier deals in the pipeline.


Many carriers are thinking about building their own DVB-H nets. But in the European home of the technology, it's TV broadcasters, not cellular carriers, that have the required spectrum (search www.eet.com for article ID: 60405616).


That has opened the door to an unlikely leader. Crown Castle International Corp., a company that manages a network of more than 10,000 wireless towers, is spinning out a subsidiary to tap its towers — and its $12 million in purchased 1,670- to 1,675-MHz spectrum — for a national mobile-TV network in the United States. "We're probably the first along in terms of getting a DVB-H network up and deploying services," said Michael Ramke, vice president of business development for Crown Castle Mobile Media.

Crown Castle describes its offering, which it will sell wholesale to cellular carriers, as the basic cable of wireless. The suite of about 10 video channels — including news, sports and entertainment — will offer QVGA-resolution (320 x 240) video running at 15 to 25 frames/second, depending on the carrier. The service will also include about 20 audio channels for various genres at 32 kbits/s, as well as some excess capacity for carrier-defined programming.


Later this year, Crown Castle plans to announce handset, carrier and content partners when it turns a technology trial that's been running for a year in Pittsburgh into a full-blown commercial trial. Full commercial service is expected to switch on by June, with "a fast path to the top 30 U.S. markets," said Ramke.


Motorola, Nokia and Samsung participated in the tech trial, and Microsoft will provide its media player and digital rights management software. Crown Castle won't say where it will first launch its mobile-TV services or how long it will take to go nationwide.

Good news, bad news
The good news for Crown Castle is that it has the spectrum and towers it needs, a complete and open standard in DVB-H, and a core group of technology partners.


On the downside, it is limited to a relatively low, 2-kW maximum transmission power on a relatively high, 1.6-GHz frequency. Higher frequencies translate to shorter propagation distances — meaning Crown Castle's services could require more towers, basestations and operating costs and could generate greater multipath-interference problems than competitors such as archrival Qualcomm's MediaFLO.


FLO, or forward-link-only, technology can transmit at up to 50 kW at its given 700-MHz frequency, though Qualcomm won't say what average-power levels it will actually use. The company plans to employ two or three high-power transmitters in urban areas, where Crown Castle will need to use 20 to 30, implying lower costs of service for FLO.


In addition, FLO can pack 20 channels into 6 MHz of spectrum, compared with nine channels for DVB-H. And it has a channel change time of 1.5 seconds, vs. 5 seconds for DVB-H.


But Qualcomm's 700-MHz spectrum is encumbered by the large number of analog UHF TV stations broadcasting at or around channel 55 — crowded conditions that inhibit access to as many as 180 million users, according to Crown Castle. Qualcomm is already approaching some of those broadcasters, but it may have to wait until late 2009, and a mandated U.S. deadline for the shift to digital TV, before it can assemble a nationwide network.


Further, the FLO spec, while apparently finished, has yet to be openly reviewed and codified by the industry. Qualcomm set up the FLO Forum, with 14 industry members, in late July to review the basic physical-layer spec under a nondisclosure agreement — a process that will take less than a year. The group must then shepherd the spec through industry standards organizations, which will take at least another year.

On the business side, Crown Castle admits it lacks the funding to build a nationwide net. The company has annual revenue of about $600 million and has sustained significant losses in four of the last five years. It is in talks with equity investors for spinning out its mobile unit to pay for building its network. In addition, it has no past links to content providers, and it is not known for its marketing prowess — two hurdles for success in mobile TV.


For its part, Qualcomm said it may take on investors. But with annual revenue growing toward $5 billion and a track record of profits, the company may have the ability to finance a net buildout alone and then spin out the unit in a public offering. And as the licensor of core CDMA technology, Qualcomm likely has deeper carrier partnerships than Crown Castle.


Qualcomm also has been developing relationships with content companies for several years, courting them to develop applications and services for its Brew handset software platform. And it has earned some consumer marketing expertise promoting Brew.


The MediaFLO service is expected to launch in October 2006, with about 10 QVGA video channels running at up to 30 frames/s, at a monthly subscriber cost of less than $20. FLO can deliver a maximum of 20 video channels and 1,000 minutes of video clips.

Now playing
As Crown Castle and Qualcomm race to build their new multicast nets, startup Idetic is already gaining traction for a service delivering television over today's cellular networks.


The startup's MobiTV service can deliver streaming video over links as narrow as 18 kbits/s. Idetic has relationships with more than 25 content providers and offers services to any handset that supports Java or Brew on the nets of at least four cellular carriers. It is also starting a trial to deliver standard-definition video over Wi-Fi, so future coffee drinkers at Starbucks might watch TV on their laptops.


"By the time Crown Castle and Qualcomm launch their services, we expect to have 1 million to 2 million subscribers and the ability to use whatever infrastructure wins out," said Idetic's Alvelda.


The company started in 1999 as a provider of C++ software for optimizing delivery of video over cellular, but carriers were focused only on text data at the time. "So we set off to create the problem we had solved by delivering streaming broadcast video," said Alvelda, who hired Hollywood executives to strike content deals and put the company in the services business.


Quality and capacity are the twin Achilles' heels for overlay services like MobiTV. It started broadcasting at just 2 frames/s. It gets up to 15 frames/s on today's top phones but needs networks to upgrade to 3G technology to approach the 25-frame/s rates of DVB-H and FLO.


The one-to-one switched nature of the cellular network also puts a capacity limit on overlay services such as MobiTV. The service will scale to handle 20 million subscribers nationwide, Alvelda said, but if mobile TV goes mass-market, it will require a separate multicast network in cities.


In the end, carriers think they will need coordinated video capabilities linking their existing cellular and new multicast TV nets. But their first step is gauging real demand for phones that combine links to broadcast and cellular nets.


British carrier mmO2 plc is starting a trial this month that offers 16 channels over DVB-H to 350 users in Oxford, U.K. The trial runs through February, but "we'll have a good idea of demand by the end of October," said R&D vice president Mike Short.