Wireless could jump-start market recovery: but only if 3G gets its act together.

Will this year finally be the year of third-generation mobile phone technology? That's the question that users, OEMs and analysts are asking as the industry seeks to recover from a dismal 2002.

But while hope springs eternal, the realities of the market seem to indicate that 2003 will primarily be a transition year, with vendors rolling out--but not yet offering-high-speed services to test the waters, at least in the United States. As we have noted in the past, though, reviewing the mobile phone situation in Europe can be helpful when trying to make sense of probable deployment schedules and likely acceptance rates here.

Nokia, the biggest presence in the European mobile market and the largest mobile phone OEM in the world, continues to be the industry bellwether. The company believes the total market volume for handsets reached 400 million units in 2002, and in 2003 market volume will grow 10% or slightly more. In December, at the company's biannual strategy meeting for investors in Irving, Texas, company chairman and CEO Jorma Ollila indicated that he expects growth in 2003 will be driven by a combination of subscriber growth and a stabilizing replacement cycle.

According to Ollila, currently slightly more than a quarter of global subscribers upgrade their handsets annually, which means an average replacement cycle of about 2.5 years. This cycle has lengthened over the last 18 months, but Nokia feels it now appears to be stabilizing. The company projects that the mobile subscriber market will grow from more than 1.1 billion at the end of 2002 to approximately 1.5 billion in 2005. In 2003, Nokia expects to ship 50-100 million color handsets, all with Multimedia Messaging Service (MMS), Java and browser support. Of these, approximately 10 million will be based on Nokia's Series 60 user interface--based on the Symbian OS.

How well is Nokia positioned for next-gen wireless? According to some industry observers, the company's significant lead in the hardware market will give it a competitive edge on the software side as new mobile applications begin to arrive. Symbian and its open source code have been licensed by the top six mobile phone manufacturers, who together account for 80% of the handheld market, according to recent research by InfoTech Trends. According to the firm Microsoft, with the introduction of the Orange SPV--which uses a scaled down, proprietary version of Windows--pulls a distant second.

But when will 3G be implemented? A number of US carriers have committed to rolling out advanced services this year, although most are likely to support relatively low "2.5G" speeds of 100Kbps or less, at least until 2004. "AT&T Wireless, Cingular and T-Mobile have announced plans to implement GSMIGPRS/EDGE [Enhanced Data rate for GSM Evolution] upgrades within the existing cellular frequencies, using Compact EDGE," said Ray Jodoin, director of wireless a research at Cahners' InStatIMDR. The good news, however, is that many US carriers have made the transition to CDMA2000 lx relatively painlessly. "The evolution of a cdmaOne network to CDMA lx is relatively simple since very few new elements are required," said Jodoin.

But things on the GSM front in Europe, particularly where 3G is concerned, are a mess. For carriers, there is no easy (read: cost-effective) path from GSM to 3G. "Europe's 3G effort is in deep trouble," agreed Ira Brodsky, president of Datacomm Research, a consultancy that advises telecoms and OEMs. "They decided to go for broke with new operators (a.k.a. Universal Mobile Telephone System operators), in new spectrum (2.1GHz), using a new technology (Wideband CDMA). One could argue that the US's fragmented 2G standards allowed Europe to jump way ahead. But now it's clear that Europe's one-standard-fits-all approach has come back to haunt them."

In a new report titled "UMTS at the Crossroads: Strategies for Success," Datacomm contends that because Europe chose GSM as its universal standard, and because GSM networks are so well-entrenched, existing carriers are now between a rock and a hard place: They can't go back, but going forward is tremendously expensive.

"Most 3G operators can't afford to build nationwide networks, having spent too much on licenses, and because WCDMA infrastructure is very expensive," Brodsky said. "So Europe's 3G operators need dual-mode phones so they can build out their networks in major cities only. But dualmode W-CDMA/GSM handsets are going to require years of refinement before they can compete with standard GSM handsets. They will be bigger, more expensive, bum up batteries faster, and must make it through the gauntlet of government certifications."

Razor's EDGE

What can US carriers learn from the situation in Europe? And who is leading the race to 3G in the United States? While the U.S. market may be fragmented from a technological standpoint, for the most part vendors are at least able to afford 3G upgrades to their networks. Cingular Wireless is now testing a software upgrade for EDGE, its 3G technology. (As part of its current GSM deployment-still in only half of its coverage area- radios in GSM overlay markets are already EDGE hardware capable.)

By the end of 2003, according to the company, all of Cingular's major markets will be both hardware and software enabled for 3G. EDGE is expected to at least double GPRS data rates, for peak throughput of 470Kbps. Users can expect average throughput speeds of 100-150Kbps when the network is fully loaded.

But the rub is that EDGE is a very new technology and basically untested in widespread commercial applications. "EDGE is a big unknown right now," said Datacomm's Brodsky. "A couple of years ago, it looked like EDGE was a non-starter. All of the development effort went into GPRS and W-CDMA. Now, only because W-CDMA is in trouble, EDGE is being revived."

Brodsky said some key issues about EDGE have yet to be answered. "Cingular plans to deploy EDGE in the 800MHz band. They may be the only carrier in the world to do that. How many handset makers will develop and sell a handset that can't be used anywhere else? How will that affect price?" Further, it's possible that EDGE may require changes to cell site spacing. "If EDGE requires cell sites to be closer together, it could get very expensive," contended Brodsky. "EDGE is just a faster version of GSM, and that requires either better signal strength or better receivers at both ends."

Other vendors have chosen the CDMA upgrade path over GSM and EDGE. US Cellular announced in December that it signed a $100 million agreement with Lucent Technologies for the deployment of Lucent's 3G CDMA 1xRTT infrastructure equipment and professional services. The network equipment will support US Cellular's conversion to CDMA 1xRTT, which began in October and is scheduled for completion in 2004. In January of 2002, Verizon wireless launched its Express Network, with speeds of up to 144Kbps available (though actually throughput is closer to 60Kbps), while Sprint launched its CDMA 1xRTT service, called PCS Vision, in August. AT&T and DoCoMo say their 3G network will be operational in San Francisco, Seattle, Dallas and San Diego in late 2004. (Because the ITU does not specify end user throughput for 3G, there is no firm definition of what constitutes "third-generation" speeds, though experts define 3G technologies as those with a link rate of 144Kbps to about 2Mbps.)

More Ream To Roam

Late last year, in a widely praised ruling, the Federal Communications Commission finally announced that it will release more spectrum space for wireless devices, for the expressed purpose of supporting 3G services. The FCC allocated 90 megahertz of spectrum space (two contiguous 45 megahertz frequency bands) for fixed and mobile wireless services. FCC Chairman Powell said that the release of the spectrum space, which should be available in 2004 or 2005, was key to helping the ailing telecom sector.

"I previously identified new revenue sources and new services as among the key steps to recovery in the telecommunications sector," Powell said. "By our action today, we will make available spectrum resources that carriers and the consuming public demand--a major step in creating an environment hospitable to the introduction of new and innovative products and services. Access to new spectrum is not a cure for today's financially ailing wireless industry, but it is a key precondition to the long term health of the industry."

Chip Problems

But while the release of the spectrum space was welcomed, it does little to change the immediate financial picture for wireless component OEMs, who continue to struggle. "2002 could be considered a transition year for cellular base-station components, said Allen Nogee, a senior analyst at InStat/MDR. "Both ASIC and DSP designs had their supporters, and new designs were created using both technologies. Overall, most new development was in W-CDMA base stations, but many carriers have delayed their W-CDMA deployments, so demand was not as high for these components as initially anticipated." (W-CDMA is the 3G path from GSM, and is currently being eclipsed by CDMA 1x/2000.)

But even more worrisome for chipmakers is the fact that their technology is now so advanced that it may be putting hardware upgrades off indefinitely. "Many technology upgrades of present base stations and the majority of upgrades in the future will not require new hardware," Nogee said. "In fact, most upgrades of CDMA, GSM, and W-CDMA base stations can be performed with software, and sometimes firmware upgrades."

Another challenge, of course, is the fragmented nature of the U.S. cellular telephone market. With some carriers moving to GSMIEDGE and others to CDMA2000, the pro-spect of continued proprietary networks--even with 3G services--is very real. Organizations like the Open Mobile Alliance are attempting to create open development platforms and network infrastructures to solve these problems.

But the reality is that, today, GSM and CDMA networks can only communicate via the GSM1x overlay. According to Qualcomm, the main force behind CDMA and the creator of GSM 1x, GSM1x enables coexistence of CDMA2000 1x radio access networks with GSM-MAP core service networks in existing or new spectrum. GSM1x offers peak data rates of 307Kbps per sector in a 1.25MHz channel and peak data rates up to 2.4Mbps with 1xEV-DO (High Data Rate), allowing GSM carriers to offer 3G services without major new equipment investments. GSM1x also allows global roaming. (1xEVDO is currently commercially available only from Monet Mobile Networks in the northern midwestern states.) But CDMA-based carriers in the U.S. may balk at upgrading their networks just so they can be used by their competitors' customers.

Another point that deserves consideration is what, exactly, customers are looking for in a high-speed mobile network. Based on the current crop of phones and the carriers' advertising campaigns, one might assume that users are demanding multimedia, Internet gaming, and digital photography from their handsets. But some analysts believe that carriers seeking to address such perceived demands may be in for a rude awakening. "The mobile industry needs to refocus its early 3G efforts," said recent research from Ovum, a telecom consultancy. "Mobile operators rolling out 3G networks need to forget about bandwidth-hungry applications for the few, and instead concentrate on narrow-band applications--including voice--for the many."

Ovum contends that creating a 3G network to provide multimedia applications to a large number of people over a wide coverage area requires huge investment, which, in the current economic climate at least, seems destined to offer a fairly poor return. "Handing back the licenses and pretending 3G was just a bad dream is not the answer," said Julian Hewett, Chief Analyst with Ovum. "Instead, operators must roll out 3G gradually in traffic hotspots, such as capital cities, where many 2G networks are already feeling the strain."

Such a strategy, Hewett feels, will result in much lower levels of capital expenditure--which should mitigate some of the financial pressure operators are currently under. And once it's in place, Hewett says, 3G technology will provide a much more cost-efficient network. Voice transmission could cost 30% less than 2G, and data transmission 80% less than 2.5G technologies like GPRS.

Still, the moderate 3G coverage that exists today has shown only modest subscriber growth, mostly because such services cost $100 per month or more. At about half this cost, these services would be sure to find a broad audience, since most users pay $30-$60 per month under their current service plans. Certainly, better coverage and faster throughput, plus added features, are worth a price premium. But few customers--with the possible exception of those with particularly heavy usage patterns--are willing to double their monthly bills. Whether carriers can make money on their 3G investments with only moderate price increases for their customers remains an open question today.

www.att.com

www.cingular.com

www.nokia.com

www.sprint.com

www.uscc.com

www.verizon.com