Atlantic Crossing

Last month, Qualcomm announced its first carrier customer for its binary runtime environment for wireless, or BREW, platform in Western Europe. The deal with Nordisk Mobiltelefon in Sweden at first glance looked like just the deal Qualcomm has been waiting for — its first BREW rollout with an operator in Western Europe, a market that seemed almost impenetrable just a year or two ago. Its new strategy to target the GSM community — and more specifically Europe — seemed to be working.

Except for one thing: Nordisk isn't a GSM operator. The rural Swedish carrier is one of the new breed of CDMA carriers offering service in the 450 MHz band, using the high propagation spectrum to reach under-served markets. Score another in a long succession of CDMA carrier BREW wins, but still no tick in the GSM column.

In the last year, Qualcomm has redoubled its efforts to crack the GSM market. It bought European content distribution platform developer Elata this summer and user interface developer Trigenix last fall, acquiring both firms' GSM-centric technology and their connections in the European carrier market. With the Elata deal, Qualcomm inherited its first major GSM customers, Hutchinson Telecom's 3 and Orange France, but those carriers are still using the UMTS content management platform developed by Elata. Qualcomm may have its first insider track into the European market, but it's still a long way from selling them BREW.

Qualcomm is doing more than just trying to buy its way into Europe. The vendor revealed in recent interviews with Wireless Review that it plans to radically revise the BREW platform this fall, abandoning the end-to-end, all-or-nothing approach that, while popular with its CDMA customers, is anathema to how GSM providers operate. Instead, Qualcomm is making BREW modular, untethering its content delivery and management platform from its content service and library, and supporting other content formats aside from its own proprietary runtime environment. It is even offering support for third-party content. Basically, Qualcomm is altering the fundamental business model of BREW while still keeping the core technology intact, said Gina Lombardi, senior vice president of marketing and product management for Qualcomm Internet Services.

“Whichever pieces of the BREW puzzle the carrier wants, we can deliver it,” Lombardi said. “We have an array of options — we feel we have something for every carrier.”

BREW has always had two basic components: a content management and delivery system — which handles the distribution, downloading and payment of applications — and the applications themselves, which are developed, loaded and run entirely within the BREW framework. In others words, BREW is both a way to deliver content and a library of price-negotiated content unto itself. The two components have always been seemingly inextricably linked. In fact, much of the appeal of BREW to its proponents is its top-to-bottom nature.

“The big advantage of BREW is that it gives you an ecosphere,” said Roger Entner, Ovum wireless analyst. “It's very clear how all of the payments go down the value chain. Java and Symbian don't have that kind of clarity.”

But that ecosphere comes with limitations. BREW's operating system is embedded directly into the phone's chipset, meaning all BREW carriers had to have Qualcomm-powered phones. While not a big problem in the CDMA world, where Qualcomm produces the majority of the chipsets, the company only produces a fraction of the chipsets embedded in GSM phones. So far, no GSM handset vendor has announced plans to incorporate BREW into its wafer designs. Furthermore, the GSM community has invested billions in building a formidable Java-and Symbian-powered content apparatus. To chuck it all in favor of a proprietary technology developed by Qualcomm would take an enormous effort. Of the content itself, BREW has built an intimidating developers program driven by the popularity of its service worldwide, but while most BREW applications have Java equivalents, the reverse can't be said. And, the Java development community far exceeds that of BREW.

All of these factors forced Qualcomm to rethink BREW, ensuring that its end-to-end functionality was still available to carriers that wanted it but making it open for those that didn't, Lombardi said. The obvious choice was to make it modular, allowing carriers to use any permutation of BREW possible. The new platform would allow carriers merely to use BREW as a management platform to track and bill for their content. As BREW gets pushed further into the network, it could take over the downloading and installation of Java and Symbian applications on the phone. If BREW is pushed all the way to the handset, its retooled Trigenix user interface would run the applications themselves, allowing access to BREW's extensive content library. That would be the ultimate step for Qualcomm, but Lombardi said it isn't counting on carriers wholeheartedly embracing BREW applications — at least not yet. Qualcomm is building support into the new BREW/Trigenix UI that allows it to run both Symbian and Java apps along with BREW content. Or if a carrier isn't comfortable with the BREW client calling all the shots, it can relinquish control to the Symbian or Microsoft OS or a middleware such as Nokia's Series 60 to run specific apps, Lombardi said.

While Qualcomm's plans may seem entirely new, there is precedence. One of Qualcomm's biggest clients, Korea's KTF, actually uses a modified version of BREW in its networks, using the delivery platform to provision and manage content but populating its content portal with Java applications. Qualcomm isn't just giving the modular approach lip service, it's already executing the strategy with its established CDMA customers, Lombardi said.

Still, despite the philosophical shift, Qualcomm has an uphill battle to fight. There is still a lot of lingering distrust for Qualcomm left over from the air interface wars, which many of Qualcomm's critics have accused it of pursuing with religious zeal. And while the GSM community may admit Qualcomm into Europe as a player in UMTS silicon technology, letting Qualcomm run its content services is another story.

“Don't forget, Qualcomm has been trying to get into Europe for a long time, and they haven't succeeded,” said Andrew Cole, chief telecom analyst with A.T. Kearney. “There's an embedded bias against them.”

Lombardi admits that this reputation is Qualcomm's worst enemy, but she claimed Qualcomm has made progress in improving perceptions in Europe. Not only has the inclusion of the firm's core CDMA technology in European 3G standards increased its involvement with the community, but it has been reaching out to GSM carriers through its European acquisitions. Most important, Lombardi said, Qualcomm is letting the technology speak for itself. Customer CDMA 1X EV-DO networks running BREW in Asia, Brazil and the U.S. are impressing the old guard operators in Europe, and they're becoming convinced that Qualcomm really can them an open solution, Lombardi said.

“Three or four years ago, we weren't so welcome in Europe,” said Lombardi. “We're a lot more welcome there now.”

Qualcomm isn't the only big vendor to try to crack the European content market. Last year Nokia launched Preminet, a similar content management and distribution network, complete with an application catalog from its sizable Forum Nokia development community. So far, the service has yet to take off. Nokia has publicly announced only two customers for the Preminet — SmartCom Chile and India Games in South Asia — though Germany's T-Mobile has been evaluating the service. Preminet suffered from Nokia's top-heavy approach to the market, said Anil Malhotra, vice president of alliances and marketing for Bango, a third-party content distributor.

It was difficult for content developers to get into Preminet's content library, and although the service appeared to work exceptionally on Nokia's own phones, its support for other vendors' devices was less than stellar, Malhotra said. The same problems hindering Preminet could entangle BREW. While Qualcomm has demonstrated seamless operation of its service over an end-to-end BREW deck, its support for Java and Symbian apps is still unproven. Furthermore, if it ventures beyond BREW-enabled handsets into other vendors' operating systems and user interfaces, the complexity could hamper any vendor, not just Qualcomm.

“Qualcomm philosophically is trying to avoid those traps,” Malhotra said. “Whether practically they succeed remains to be seen.”

Creating a modular platform isn't as easy as it sounds, Malhotra said, especially if that platform was originally designed to be an end-to-end solution. He pointed to Apple and its attempts to break the Apple platform into its various components so it could license its platform.

“They found it was quite hard to disassemble those components to license them off,” Malhotra said. “It was not a challenge Apple was able to rise to.”

Also posing a challenge is the very crowded European market, with vendors, platform providers and content developers all vying for pieces of the carrier portal. Openwave and Sun Microsystems are new entrants into the content distribution arena. Motorola is expected to follow Nokia with its own content platform. The independent platform providers have also carved out names for themselves. Bango builds its business on off-deck content that isn't carried on its customers' deck, providing payment, download and management services for reams of third-party applications. Then there is End2End, a company that not only manages and delivers content but ultimately runs a carrier's portal, offering a fully outsourced solution down to the individual applications residing on the deck.

According to Atte Miettinen, chief marketing officer for End2End, the advantage in Europe Qualcomm hopes to gain through its acquisitions of Trigenix and Elata may be fleeting. Customers of those companies are fleeing to other providers, Miettinen claims. In fact, just a month after Qualcomm announced its purchase of Elata, End2End lured away one of Elata's customers, Portugal's Optimus. Regardless of how much Qualcomm maintains its openness, the fact is that BREW is still the kind of proprietarily developed technology from which the GSM world tends to cringe.

“Things are very standardized here — Europe is a very open market,” Miettinen said. “For Qualcomm to bring BREW as a proprietary technology here will be very difficult.”

Still, even Europe occasionally dallies in proprietary systems. Miettinen pointed out that NTT DoCoMo's i-mode has enjoyed some success in the limited markets in which it's been deployed. While a huge network offering, i-mode is still essentially a walled garden. Systems like i-mode, however, tend to be the exception to the rule. If Qualcomm plans to succeed, it will have to make good on its commitment to an open platform, Miettinen said.

Perhaps the biggest impediment for Qualcomm now is the relatively few number of BREW handsets in the market. BREW is embedded into the W-CDMA chipsets Qualcomm sells, meaning not only are the handsets made with other vendors chipsets off limits to BREW but so are the millions of 2.5G handsets Qualcomm doesn't support.

Lombardi, however, said those factors are built into the QIS's business strategy. The Elata acquisition gives it the basic platform to support Java and Symbian delivery, and ultimately the BREW delivery platform can feed content to non-BREW-enabled handsets. To bring BREW's full implementation to customers, however, it will have to get BREW phones in their hands. Last year 11 million UMTS handsets shipped to Europe with Qualcomm chips inside. This year Qualcomm estimates that number will be 20 million. Furthermore, Qualcomm is actively looking for other UMTS vendors to license its BREW technology, no longer limiting a BREW application to a Qualcomm-powered phone, Lombardi said.

While getting Nokia to license BREW for its own UMTS handset line appears a tough sell, a major European GSM carrier adopting BREW could change that perspective quickly, giving Qualcomm the stamp of legitimacy and the momentum it wants to propel not only the BREW service into Europe but also adoption of BREW handset technology. But landing that first GSM customer will take some time, Lombardi conceded.

“The bigger guys we've been talking to for a year and a half,” Lombardi said. “You thought the U.S. was the slow. These guys are being very careful.”

Qualcomm is being cautious on its timeline, but that hasn't stopped the rumor mill from churning. According to rumors circulating among European vendors, Qualcomm has already recruited its first GSM carrier and is preparing for a big announcement later this year. Perhaps Nordisk was the object of those rumors, but according to Bango's Malhotra, there's word of a much bigger GSM deal in the wind — the deal that will finally give Qualcomm the GSM approval it craves. Qualcomm predictably said it wouldn't comment on rumors. But it's certainly not killing the buzz.

“If I was Qualcomm, I'd stir that rumor up myself,” Malhotra said.

Current BREW Operators
(listings are U.S. unless otherwise noted)

Alaska Communications
Alltel
Appalachian Wireless
Bermuda Digital Communications	Bermuda
Bluegrass
Cellcom
Cellular One
Cellular South
Centennial	Puerto Rico
China Unicom	China
Cricket
Eloqui Wireless
Golden State Cellular
GUAMCELL Communications
Hutch	Thailand
Illinois Valley Cellular
Iusacell	Mexico
KDDI	Japan
KTF	Korea
Metro PCS
Midwest Wireless
Movilnet	Venezuela
Nordisk Mobiltelefon	Sweden
NTELOS
Pelephone Comm. Ltd.	Israel
Pioneer
RCC Wireless
Reliance Infocomm	India
Tata Indicom	India
Telstra	Australia
U.S. Cellular
Verizon Wireless
VZI-Dominican Republic
VZI-Puerto Rico
Vivo	Brazil
Zapp	Romania
Source: Qualcomm