3 wishes
The 3G guessing game. We look into big bandwidth and big questions.
Terrence P Lynch, Contributing Editor

Potential applications for this so-called 3G (third generation) cellular technology include enticing concepts such as handheld videophones, mobile Internet multimedia, and fully capable mobile offices. But reaching the data rates necessary for those data-based services will require billions of dollars in infrastructure investments. Undaunted, system operators worldwide are already touting wireless data services. The question is, will enough people be listening to make the investments worthwhile?

Kristoffel Mulier

The advent of 3G cellular service divides itself into two parts: a rundown of the various standards and speculation on what the new capabilities will bring to everyday life. But there's also the question of alternatives to the new technology. And companies are doing a good deal of bet-hedging as developers struggle to shape the future of wireless communications.

Standards scorecard

When the International Telecommunications Union adopted standards called IMT 2000 (International Mobile Telecommunications in the year 2000), it recognized that the disparate cellular phone standards deployed worldwide could not be reconciled. Operating frequencies differ from region to region, and the installed equipment base couldn't simply be scrapped without economic chaos.

Instead, the ITU agreed to map out separate pathways for each of the major competing technologies toward a similar packet-switched communications protocol called CDMA (code division multiple access). Different versions of CDMA would allow backward compatibility with legacy equipment, but the new equipment would remain deaf to other versions of the technology. Nevertheless, IMT 2000 called for a vast increase in data rates over the poky (9.6 kbit/sec) rate of many existing systems—144 kbits/sec for handsets moving at automotive speeds, 384 kbits/sec at pedestrian speeds, and up to 2 Mbits/sec indoors.

In short, IMT 2000 retained the status quo while improving data-exchange rates and easing the design burden for component manufacturers—single CDMA-compatible ASIC cores could be used for all flavors of 3G systems, with only relatively minor software tweaking for each system.

"3G is an evolution, not a revolution," says Mike Schoonover, strategic marketing manager for the wireless communications unit at Texas Instruments. Each of the competing systems has a series of technologies to deploy in order to economically and gradually reach the 3G standard.

Littered as they are with acronyms and trademarks, the technical migration paths to 3G (see Figure 1) can be difficult for experts—not to mention the uninitiated—to follow. So take a deep breath.

The circuit-switched, TDMA (time division multiple access) system known as GSM (originally groupe speciale mobile but now the global system for mobile communications) is in the process of changing over to GPRS (general packet radio services). The switchover is a major undertaking, involving as it does some two-thirds of the world's operators.

GPRS improves transmission rates to 56 kbits/sec and will enable a further transition to EDGE (enhanced data rates for GSM evolution) starting in the second half of next year. EDGE provides peak rates of 384 kbits/sec, enough for low-frame-rate video applications and high-speed Internet access. However, full 3G compliance for today's GSM will require a change in the so-called air interface, or operating frequency, plus the deployment of still a third protocol, wideband CDMA, sometime thereafter but before 2003. Wideband CDMA is also known as Direct Spread technology and sometimes as UMT (universal mobile telephone services).

Other TDMA networks—IS-136 in the United States and PDC (personal digital cellular) in Japan—have migration paths similar to GSM.

Exhale. Now take another deep breath.

The 2G circuit-switched CDMA networks (also known as IS-95 and cdmaOne) now deployed around the world have already added packet-data capabilities by switching over to a protocol called IS-95A. Many operators are currently installing IS-95B, a streamlined version of the protocol, which enables data speeds to 64 kbits/sec.

As early as this month, operators in Korea will begin to deploy 3G technologies called CDMA-2000 or Multicarrier. The initially deployed version of CDMA-2000, called 1x Multicarrier, runs on a 1.25-MHz channel width and will allow transactions at up to 144 kbits/sec. A version called 3x Multicarrier (3.75- or 5-MHz channel width) will follow, with HDR (high data rate) versions after that capable of delivering speeds to 2.4 Mbits/sec.

It is now safe to resume your normal breathing.

Although the intermediate steps to 3G described above are sometimes referred to as "2.5G," no formal 2.5G specification exists. That companies feel the need to coin such an ungrammatical term reflects the incremental, though by no means plodding, nature of 3G adoption.

The reason for the stepwise approach? Economics, of course. With a worldwide cell-phone customer base approaching 1 billion people, system operators face a massive job swapping one technology for another.

Varying impact

Infrastructure costs vary with the technology in question, naturally. According to Bharat Shah, director of business development at Qualcomm, operators who've already made the switch from circuit-switched technology to IS-95B should have little trouble making the change to 3G. "From the air-interface perspective, supporting these new channels, there's only a minor upgrade, a channel card upgrade," he says. "It's the same RF spectrum, so nothing changes."

Lars Nilsson

One variable in the 3G cost equation is, of course, the cost of developing new handsets to make use of the new infrastructure equipment. The handset issue is especially important, says TI's Schoonover, because to make money on the new services 3G offers, operators will have to provide handsets with reasonable cost and with accustomed performance. "Today, you have phones that have two or three hours of talk time," he says. "If you add features and there's only 30 minutes [of talk time], people aren't going to buy it."

Fortunately, the 3G protocols themselves assist designers in the power consumption dilemma. "[3G] co-processing methodologies consume less power because transmission time is optimized over 2G," explains Luis Pineda, vice president of product management at Qualcomm CDMA Technologies. "The modulation schemes and the coding schemes [are] more efficient, giving better capacity of the spectrum." Steady engineering improvements like smaller component geometries and lower-voltage operation should also reduce anxieties over handset performance.

TINY DESKTOP: Roughly the size of a twin-CD jewel box, Ericsson’s pocketable 3G-device mockup includes a foldable keyboard for email and other computer-like functions.

Kristoffel Mulier, Sirius' founder and sales and marketing director, says that his company's 3G DSP, called CDMAx, can be hardware-configured to meet the requirements of the various 3G standards as well as the various roles a 3G device must play. "With a 3G mobile handset," he says, "you'll sometimes use it only for voice at a minimum data rate. A fraction of second later, you'll use it for a video transmission....In one mode you're making a phone call, the next moment, during an emergency, you push a button and the hardware is reconfigured so the demodulator is a GPS demodulator, [which] calculates your position, then [the hardware] switch[es] again to UMTS mode."

CDMAx, a software-reconfigurable IP (intellectual property) core, as well as 3G chipsets like TI's C6X and 5000-platform DSPs are already in production worldwide.

Future shapes

That 3G wireless systems will arrive on time with the promised data-handling prowess seems to be a given. What customers will do with these new capabilities remains open to debate. In the early 1980s, could anyone have foreseen how the PC would be embraced and changed by the needs of everyday consumers? 3G wireless faces a similar uncertainty, but also a similar hope.

In the absence of a killer application, few people today need a 3G phone. Industry leaders expect that to change.

As TI's Schoonover puts it, "In 3G, we think that the open architecture of our DSPs will allow for unforeseen applications in the future." Ericsson's Nilsson agrees. "The service layer is opened up to third-party developers," he says. "You're opening up the opportunities for third-party application developers to introduce new services. They won't be so constrained as they have been before." The industry is betting that the ingenuity of young software developers around the world will soon make 3G-enabled devices as indispensable as 2G phones are today.

Paul E Jacobs

But clearly, the industry hopes that new data-centric services will eventually form a significant portion of 3G-system revenues. Early wireless-Web applications and NTT DoKoMo's i-Mode service in Japan, they say, show the promise that wireless data offers.

Many existing wireless Web services employ WAP (wireless application protocol). WAP is a client-server architecture that compensates for the small screens and low memory of handheld devices. With WAP, users can select a limited number of Web applications that they'd like to access, then view a truncated version of those applications remotely through their phones. Typical uses are stock quotes, restaurant locators, and movie listings.

With Japan's i-Mode service, some 10.6 million subscribers can purchase tickets for travel or entertainment, check bank balances, read and send e-mail, and access pre-selected Web sites for weather information and the like. In a switch from standard industry practice, the 9.6-kbit/sec packet-data service charges by the volume of data sent, not by the airtime used. A competing system from DDI offers WAP services with handsets featuring an interface called Phone.com. That service just passed 3 million users. With interest this high and alternative pricing schemes underway, Japan may be showing companies how to recoup their investments in the future.

Beyond the phone

Dissenters downplay the significance of the Japanese foray into the wireless Web. Jeff Morris, marketing manager at AirCard modem maker Sierra Wireless, wonders whether 3G systems will live up to their promise. "They talk about rates to 2 Mbits/sec, but there's a difference between the theoretical speed they talk about and [actual performance]," he says. "They get those high data rates by assigning for short periods of time half a dozen voice channels to that user. Are the carriers really going to allow that? They don't know."

Typical road warriors, Morris says, often have laptop computers along with PDAs and cell phones, and they often use two or three of these devices simultaneously. The inference is that a Web-enabled phone, with its necessarily small screen, will not be able to meet the needs of business people on the go. And business people are expected to be the early adopters of any new wireless technology.

MAGIC WAND: Incorporating streaming video, Internet access, and digital telephony on a multifunction touchscreen, this mockup from Sirius shows one vision of a 3G handset.

Mulier notes that i-Mode's usage patterns have startled even the system's developers. Only about 10 percent of the traffic concerns business applications like stock trading. Fully half the traffic involves less weighty data such as horoscopes and music files, he says. "Where a mobile phone today is a tool for management, the 3G market will be dominated by teenagers," Mulier predicts. "The mobile phone will be a gadget. We're entering into the toy business."

Whether 3G cell phones are successful in the hands of business people or teenagers may be beside the point. "The thing that we think is going to drive 3G is data," says Paul E Jacobs, Qualcomm's executive vice president. "And the key issue there is to get applications that people really care about."

Jacobs points to his company's recent announcement of a joint venture with Ford—a new company called Wingcast, which will provide in-vehicle 3G CDMA links. The Wingcast system will enable communications, information downloads, security services, and remote services like keyless entry and accident response. "[It's] about growing the market and getting more devices per person," Jacobs says. "Clearly the voice market is going to continue to grow the way it is, but what's going to drive things is the new data applications."

The biggest news about third-generation cell-phone technology may end up being that it's not about cell phones at all.

Author information

Contributing Editor Terrence P Lynch is a freelance technical writer based in Newton, MA. His work has appeared in EDN, Design News, Security Technology & Design, and numerous Web publications.