3G Migration: Waiting for the Wave

"We've been sitting around at the beach with our surf boards, and we're just waiting for that big wave," said Umesh Amin, AT&T Wireless Services director of new technology and planning. "We know that data is coming. We have done extensive soul searching, and we have clearly identified the need for high-speed data."

Nonetheless, industry progress toward 3G has moved slowly. That's no wonder. Carriers have been focused on maintaining and building out their current networks as subscribers grow to between 250 and 300 million users worldwide.

"Carriers have been quite busy with the work at hand," said Bukasa Tshilombo, Dataquest Mobile Communications Infrastructure and Services division senior analyst.

That slow period may be picking up. Ericsson and Qualcomm recently agreed to jointly support one single world CDMA standard, and the ITU is scheduled to convene in June to make its recommendations on global standards.

But there is still plenty of speculation surrounding 3G. GSM and TDMA carriers argue that multiple standards worldwide are the best approach for the industry and consumers.

"Let the marketplace decide," said Mike Houghton, GSM Alliance spokesperson.

Meanwhile, CDMA carriers favor a converged worldwide standard.

"If there is not a worldwide CDMA standard, we believe that 3G will be less accepted by the customer because of confusion," said Howie Waterman, Bell Atlantic Mobile (BAM) spokesperson. "There will be a number of lost benefits of standardization."

CDMA carriers claim that a single standard is key for interoperability, global roaming and economies of scale.

But at least one thing is clear -- whatever standards are adopted must be backward compatible with 2G networks. As the final decisions and touches are made on these standards, a few carriers are beginning to make plans for migrating to 2G+ and 3G networks.

THE SHIFTTechnically, it is not yet clear how difficult or easy it will be for carriers to make the 3G transition. Depending on standard decisions, it could be a simple upgrade to an existing core network, and 3G may be inexpensive, especially for PCS carriers that still are building networks. At least, that's what vendors are promising.

"It's some very minor changes to the networks," said Keith Shank, Ericsson director of strategic marketing & business development. "There is some investment involved but not a major forklift operation."

If, on the other hand, you need to build entirely new infrastructure to offer high-speed services, then you probably are looking at a much greater capital investment. If you must build a whole new base-station grid, then you will face onerous site-acquisition, tower anda host of other costs. But that will not be the case for most carriers.

"If any 3G service is going to be economically viable, it would be hard to imagine that they could launch something that could cost more than the current network, unless there were some dramatic capacity or data-rate differences," said Brian Bolliger, Lucent Technologies director of wireless market strategy.

The specific cost of upgrading to 3G depends on each carrier's starting point. AT&T Wireless' Amin points out that the goal is not just high-speed data, but high-speed data delivered over efficient packet networks.

The total cost of 3G upgrades will depend largely on whether you need more spectrum. If a CDMA carrier can use existing spectrum, upgrading is a matter of channel-card plug-ins with the same base stations, power amplifiers and filters. Similarly, if you are adding EDGE and have enough spectrum -- and the base-station platform can accommodate EDGE and TDMA radio architecturally -- then the upgrade path should not be steep.

Most TDMA and GSM networks will make an evolutionary transition into EDGE via new software loads and equipment additions such as some packet-switching nodes or routers.

TDMA carriers also will need a new radio in each base station. The newest TDMA standard will come in two flavors -- a 200kHz version similar to GSM and a concatenated version for 2MHz. There also is the possibility of migrating to 2G+ based on attempts to boost the data rate using some of the current radio technology. GSM's generalized packet radio solution (GPRS) is an attempt to get a packet data-based capability on the GSM physical radio channel. That method would offer data at speeds between 64kb/s and 144kb/s (thus the 2G+ banner) using current air interfaces.

"Many of the GSM operators will use EDGE in some markets, but seem to be heading down the path of GPRS in others," Bolliger said.

CDMA carriers, on the other hand, will follow the ITU's cdma2000 guidelines. CDMA carriers are adamant about maintaining existing front-end chip rates that maintain backward compatibility with 2G. Carriers will achieve cdma2000 by adjusting modulation and introducing a new channel card that supports the new capabilities in the current mobile services.

Because it still will be a 1.25MHz radio interface, most CDMA carriers in the United States will have available spectrum in their current bands. That should make it relatively simple to reach peak data rates of up to 307kb/s with an average data rate of 64kb/s to 144kb/s.

But that may not be enough.

"Now people are getting bandwidth starved, so they want data rates higher than that," Bolliger said.

As a result, some CDMA carriers may pursue the cdma2000 5MHz option that gives code rates of up to 2Mb/s. PCS carriers, with 15MHz spectrum in most cases, will find this particularly appealing because they can dedicate part of their spectrum to the new 5MHz radio and have a fair amount of 384kb/s to 2Mb/s data service in that set of radio infrastructure.

PROCEEDING WITH CAUTIONBut carriers aren't rushing to deploy these 3G enhancements. Sprint PCS was one of the first carriers to publicize 3G plans when the company committed to year 2000 lab testing last February. Sprint PCS also demonstrated high-speed wireless data using some 3G components at CTIA's Wireless 1999 conference, and the company plans similar demonstrations with Lucent, Nortel and Qualcomm.

But for spectrum-rich carriers such as Sprint PCS, commercial deployment may be at least five, if not 10, years away. Like other PCS carriers, Sprint started building its 2G network barely two years ago.

"We have a substantial brand new investment on the ground," said Tom Murphy, Sprint PCS spokesperson. "We're just using a fraction of our spectrum today and are a long way from seeing any capacity issues arise in our network that would demand a move to 3G immediately."

Holding things up even more than technical and standards issues is that the wireless industry is not even sure how to bill for new data services. Carriers want to avoid the "all-you-can-eat" dilemma that has plagued Internet service providers. It doesn't help that wireless carriers have little experience with data.

"Data is very, very confusing," Amin said.

Although the industry will expect carriers such as AT&T Wireless and BAM to announce limited trials, almost all carriers are and will be secretive about the details. Most carriers are reluctant to discuss roll-out plans because they are still preliminary.

"We have so many people in the company currently evaluating all of these aspects of what the service mix is going to be, what the market segments are going to be, the pricing plan and what kind of terminals we will need," said AT&T's Amin.

Carriers are taking their time to carefully evaluate what markets and cell sites to upgrade first. Most carriers will target specific cell sites and shift carrier channels to handle increased traffic.

Peter Barry, director of strategic technology, said AirTouch has positively identified enterprise demand for 3G services so it will likely target areas with high business activity initially in large, international cities.

Eventually, carriers will decide where the core demands are for higher-speed services. Besides business districts, carriers also may target areas outside of city centers to cover the areas where corporate users reside.

"You would make sure that you cover those in a time frame that stimulates the market," Barry said.

As of now, carriers are not telling what kind of capital investments they are allocating to 3G mainly because vendors have yet to price the new equipment needed.

"We will initially be very cautious about spending," said AT&T's Amin.

Ultimately, carriers' deployment strategies depend on the invisible hand of the market.

"We are looking at the 2000 or 2001 time frame for initial deployment, and then, depending on market demand, we may ramp up (nationwide) in 12 to 18 months or in five years," Amin said. "It's all tied to pure and simple capitalism -- supply and demand."

THE WAITUpgrading to next-generation networks would seem like old hat to cellular carriers who just a few years ago began migrating analog networks to digital. But even these carriers are taking a slow approach to the market, mainly because the industry has changed so much since it overhauled analog networks with digital technology.

The first-generation makeover to digital 2G networks was tied to simple things such as coverage, capacity and cost. It was all about voice, and nobody estimated the penetration rates that the industry would achieve when it migrated to digital.

Now, carriers realize that more than 1 billion people may use the next-generation networks. The stakes are suddenly much higher. The ITU was not even involved in the move from 1G to 2G networks. Today, everyone recognizes the need for quality-of-service benchmarks, along with the need to integrate wireless and wireline.

"Some day in the not-too-distant future, the only device that people may have is wireless for their voice communications," Amin said. "So 3G is huge in terms of its impact."

As a result, carriers are being much more cautious of the 3G migration process than they were of 2G. A few carriers are inching forward slowly, but most still are waiting for the big wave.