The technology wars of the last decade aren't over. CDMA and GSM may have struck an uneasy balance in their respective markets' technological niches, but the debate is about to resurface — this time in another wireless technology, multimedia broadcast. The camps are basically the same: Qualcomm and its Asian CDMA partners on one side and the sizable GSM community on another. The tactics are essentially the same also: Qualcomm pitting a proprietary technology against one built out of consensus standards.

At stake is what many in the industry consider a cornerstone in the wireless data, mobile video arena — TV on the cell phone. Because of the capacity-hogging potential of such a service, the industry has envisioned entirely new networks for this new mobile TV service. Now it just has to figure out what kind of network that will be.

Qualcomm has designed, from the ground up, a multicast technology called Forward Link Only (FLO), while the traditional GSM vendors have joined with the Digital Video Broadcast (DVB) community to create a low-power derivation of the DVB-television standard called DVB-hand-held (DVB-H). The technologies aren't just different, the approaches their supporters are taking to the market are also at odds. DVB-H is being trialed by numerous carriers all over the globe using multiple vendors' equipment and separately negotiated content deals. And those trials aren't just in the GSM heartland of Europe. Crown Castle is planning a commercial trial in Pennsylvania this October.

Meanwhile, Qualcomm is playing the mother lion guiding its cub from birth to maturity entirely in-house. Its subsidiary, MediaFLO USA, is not only developing the core FLO technology and chipsets, it's acquiring spectrum, building its own network and planning a commercial launch by late next year. Qualcomm's plan is to eventually spin off MediaFLO, license FLO technology and sell FLO chipsets to other vendors like it does its other core CDMA technologies. But before that, Qualcomm isn't relying on other vendors and carriers to prove its business case. It's doing it on its own.

At first glance, the technologies look similar. The transmission modulation scheme is based on orthogonal frequency division multiplexing. They are both designed to transmit upwards of 15 frames per second of video to a Quarter VGA (QVGA) screen. And they've both emphasized power conservation, each with the stated goal of allowing for four hours of continuous viewing off a single battery charge.

But on one critical spec there is a tremendous amount of disagreement over how much the technologies differ.

“Compare the performance of the two systems on the same spectrum and you'll find FLO has either twice the coverage or twice the number of channels,” said Rob Chandhok, vice president of engineering for MediaFLO.

Qualcomm's technical white papers show that over a 6 MHz channel, FLO can fit 20 channels while DVB-H can fit only nine. DVB-H proponents dispute those exact numbers, saying that many factors from allotted bandwidth, to frequencies used determine how many video channels can be shoved into a particular swathe of spectrum.

“There are a lot of different variables determining how many channels at what resolutions can be fit into which spectrum,” said Bob Shallow, Nokia's director of rich media and music for North America. “There have been lots of very simplistic claims made about the two technologies, but when you peel back the layers of the onion you find that the technologies are very similar.”

There's also a difference between theory and implementation, Shallow pointed out — while Qualcomm has yet to put MediaFLO to work in live conditions, DVB-H is currently running in several trials worldwide. Although recent commercial trials in Finland have demonstrated multicast broadcasts of 16 simultaneous video channels with 112 kb/s streams producing 15 frames per second at sub-QVGA resolutions, this is still far below the capabilities outlined by MediaFLO. The next projected phase of testing, which elevates those 16 channels to 384 kb/s streams at full QVGA resolution, would almost match Qualcomm's projected specs for FLO, but at half the frames per second.

There have been other issues with DVB-H aside from channel capacity. Multicast video network handsets only power up their broadcast receivers at intervals in order to conserve power. While the video stream is uninterrupted during that interval, the receiver can only tune to a new channel when it's powered. That interval is much longer for DVB-H than it is for FLO, which can result in a lag of several seconds in switching channels, a problem that has always plagued digital TV — mobile or otherwise.

While DVB-H's supporters claim Qualcomm exaggerates the performance differences between DVB-H and FLO, they do readily concede that there are technical advantages to FLO by virtue of its independent development. But they also claim that the same independent development is also the biggest hindrance of the technology. By being developed in-house as a proprietary technology, MediaFLO is isolated. Without the backing of the larger wireless industry, it can't gain the traction, nor the economies of scale to succeed, said Mike Yonker, director of technology strategy for Texas Instruments' wireless division.

“With open standards, there are sometimes compromises in order to achieve industry consensus, but that industry consensus drives down costs and brings things to market,” Yonker said. “All of those small technical differences don't amount to much if you can't make a commercial case for your technology.”

Ovum analyst Roger Entner likened DVB-H and MediaFLO to a mule and a horse. By creating a consensus standard linked to an established digital broadcast technology, DVB-H's proponents have created a mule, a very strong, very sturdy creature of a technology that meets any carrier's basic requirements a multimedia broadcast service. But it's not as fast as a horse — it's no FLO, he said.

“Technically speaking, MediaFLO is a better solution,” Entner said. “They started with a blank sheet of paper and let their engineers go wild.”

But history has proven a superior technology won't always win out, Entner said. He drew up an old example: VHS won out over BetaMax despite BetaMax's obvious technological superiority. Qualcomm has challenged standards before, and no one would argue it hasn't doesn't so without success. CDMA is a major technology in the world today thanks to Qualcomm's efforts, but its reach still pales in comparison to GSM. And it's that same GSM community that Qualcomm will have to convince of its new technology merits if it doesn't want FLO to sit in the distant second spot for worldwide deployment. Therein lies Qualcomm's problem: Standardization counts for a lot among the GSM community. A widely adopted standard promises more handsets, more chipsets, more network equipment and presumably more content to distribute — all of which can achieve the economies of scale that a proprietary technology cannot.

Qualcomm, however, is trying to overcome the perception that FLO is a proprietary technology. The vendor has launched a group called the FLO Forum, which is designed to push for global standardization of the technology. Asian CDMA heavyweights Huawei, Kyocera, LG Electronics, Samsung, Sanyo, Sharp and ZTE have all signed on to the forum's membership roster, but so far none of the traditional GSM powerhouses have shown any interest. While Qualcomm developed the initial technology in private, it plans to fully develop it with the global wireless community, MediaFLO's Chandhok said. After the initial launch in the U.S., Qualcomm plans to fully license the technology from chipset to media server, and it fully expects other vendors to not just produce FLO handsets and network infrastructure but print their own silicon, Chandhok said.

DVB-H proponents maintain that making the development process more public hardly changes the fact that Qualcomm is pushing what is essentially a proprietary technology. Texas Instruments' Yonker said TI considered joining the FLO Forum when it was first proposed but found little to persuade the silicon vendor that the forum would be anything than a sounding board for Qualcomm. Instead, TI chose to funnel all of its resources into developing its Hollywood chipsets, a DVB-H integrated circuit scheduled to begin sampling in the fourth quarter.

“We weren't convinced that Qualcomm's intellectual property would result in a fully developed open standard,” Yonker said. “We felt we couldn't operate in that kind of ecosystem.”

Chandhok, however, said that the FLO's critics are merely delineating shades of gray. He pointed to other efforts such as the Open Mobile Alliance that were pushed by a single vendor (in the OMA's case, Nokia) but won general industry acceptance. Even if the FLO's initial development wasn't set by industry consensus, at a certain point the question is moot. Mobile multicast technologies aren't building within the existing network framework — they don't require integration into an existing base station or handset digital signal processor. The networks will be entirely independent with their own towers, transmitters and spectrum. Every handset will require a dedicated receiver chip regardless of which technology or standard is used.

“If there isn't a business case for FLO, don't use it,” Chandhok said. “But don't use DVB-H just because there is some perceived technical advantage to a technology developed in a public standards forum.”

While the handset vendors and chipmakers have been taking sides, the network infrastructure manufacturers have been fairly quiet about the debate. That's likely because they'll have very little to do with the launch of these networks. Instead, broadcast TV vendors like Harris Television are expected to take the lead. And so far they've remained fairly neutral.

Multicast mobile multimedia operates much the same way as the traditional TV network. In fact, it uses much of the same transmission equipment as your local NBC affiliate, including the same red-and-white broadcast towers perched on the hillside. Unlike a cellular network, which requires hundreds of cell sites to cover a market, a multicast network can operate with just two or three. Because the transmission is one way, a multicast network doesn't have to worry about a low-power cell phone having to transmit a return signal. It can effectively blanket a market with a high-powered signal from miles away.

Dave Glidden, Harris Television's director of TV strategic and business development, said his company is supporting both FLO and DVB-H rollouts. Although normally there is a risk in developing equipment for not one but two unproven technologies, Glidden said in the case of mobile multicast, the risk is far mitigated by the similarities between not only DVB-H and FLO themselves but also between the two and traditional digital TV networks, which Harris has been building for nine years.

“We're still in a very early market stage, and it's important to remember there are other mobile broadcast standards out there besides DVB-H and FLO,” Glidden said. “But there's a lot of product reuse for both technologies. We're essentially using high-power television transmitters.”

Regardless of whether vendors decide to remain neutral or take sides in the battle between DVB-H and FLO may not matter. Vendors follow the lead of their carrier customers, and that's probably why the interest in the two technologies has varied by region. The U.S. is the only region scheduled for a FLO deployment as of yet, but a lot of interest in the technology has been expressed by Asian carriers and consequently their vendor counterparts. Meanwhile Europe has been a hotbed of DVB-H activity. Chandhok, however, said that a continental split between the two technologies is far from final.

“Ultimately, these aren't decisions vendors are going to make,” Chandhok said. “It's up to the carriers, and we're getting a surprising amount of interest from carriers, especially European carriers, because of FLO's value proposition.”

So how will carriers ultimately choose which technology to pursue? In the U.S. — which is looking like it will have a mixture of DVB-H and FLO — it might come down to whom in the operator's organization is making the decision, said Ovum's Entner.

“It all depends on how strong the CTO is at a carrier,” Entner said. “A marketing or strategy guy would probably go with DVB-H. Let an engineer make the decision, and he'll go with FLO.”