Telecom's Contrarian

In Martin Cooper's view, all those high bidders for 3G licenses are in for a bitter disappointment

President Ronald Reagan's Strategic Defense Initiative failed to produce a celestial shield that could defend America from incoming nukes. But the technology that was developed with SDI grants to track intercontinental ballistic missiles is now finding a use--in wireless, hand-held gizmos for surfing the Net at high speed. Blazing new trails is a company called ArrayComm Inc. And at its helm stands one of the most colorful figures in wireless communications: Martin Cooper.

Twenty-seven years ago, at the age of 43, he stood on a street corner in Manhattan and placed the world's first phone call using a cellular handset. It was a clunky, two-pound prototype that Cooper designed at Motorola, under the skeptical watch of some of the top brass. At that time, both Motorola and AT&T were pushing the Federal Communications Commission to allocate radio spectra for car phones. But to Cooper, the cellular revolution was not about automobiles but about people using phones on the street, in stores, or wherever they wanted.

Now, he says, "history is repeating itself, with the Internet." All around the world, cellular operators and handset makers are rushing to perfect a so-called third generation of cellular technology, or "3G," which is supposed to let people on mobile phones gain access to the Net at speeds of up to two megabits per second. It was the dream of such 3G phones that fueled Britain's white-hot auctions of $35 billion worth of 3G radio spectrum to service providers such as Vodafone last month. But in Cooper's contrarian view, these bidders are heading for a bitter disappointment. "3G will be a useful voice solution, but it does little for data," he says.

FIGHTING PHYSICS? What does Cooper know that others have yet to find out? In a nutshell, he says, conventional cellular technology can't fight physics. "We have wasted billions of dollars trying to improve the air interfaces," he says, referring to the plethora of competing global cellular standards. "You can only squeeze so much information into a given channel," Cooper continues. "All the wild hype about 3G will end up yielding to reality."

Enter ArrayComm, founded in 1992 in San Jose, Calif., by French telecom entrepreneur Arnaud Saffari and Stanford University radio engineer Richard Roy. Armed with SDI grants, Roy had invented a set of algorithms for "smart antennas"--arrays of receivers and transmitters that excel at extracting faint radio signals from a cacophony of noise, ground clutter, and signal jamming. The algorithms weren't powerful enough to overcome SDI's myriad technical problems. But Saffari and Roy saw that the technology would be ideal for boosting the capacity of cellular networks, soon to be overloaded with millions of subscribers, by cutting through radio interference. What they needed was a big name in the wireless industry to run the company and persuade investors to sign on to the plan.

`COOPER'S LAW.' That big name was Cooper--whose own experience convinced him that smart antennas were the route to the future. Raised in Chicago, Cooper received an electrical-engineering degree at the Illinois Institute of Technology. After four years in the Navy, he joined Motorola in 1954, building portable police radios and eventually heading up the company's research in cellular.

Somewhere along the way, he says--laughing at his own immodesty--"I created `Cooper's Law."' It seeks to explain how use of the radio spectrum has become a trillion times more efficient since Guglielmo Marconi transmitted the first wireless signals 104 years ago. "Almost all of the improvement since Marconi is spatial, meaning the ability to reuse the spectrum over and over geographically," explains Cooper. He foresees another trillion-times improvement in the next 100 years, almost all of it spatial.

What Cooper is referring to is the fine art of painting a geographic area with very precise, low-power transmissions of energy. Cellular base stations using ArrayComm's smart antennas take signals received from the handsets of hundreds of different simultaneous users and apply high-speed signal processing to reduce the interference that each user experiences. Data are handled in the same fashion: Instead of broadcasting information across a wide area, ArrayComm's antennas create a "virtual wire" by precisely directing the data from the transmitter to the location of the receiver.

The technology has a proven track record in Japan. Starting in the early 1990s, a group of Japanese companies, including electronics giant Kyocera Corp., began deploying a low-cost digital-cellular system known as Personal Handyphone System, or PHS. Instead of powerful, costly base stations, PHS used lots of small, inexpensive antennas scattered throughout a city. Within two years, the network had 7 million subscribers. Trouble was, it was only designed to serve 2 million. Users experienced dropped calls, and PHS phones acquired a reputation as cheap alternatives to normal cell phones.

To fix the problem, PHS service providers started deploying ArrayComm software and hardware at their base stations. "Within a year," says Cooper, "the capacity of the system had increased 10 times." Analysts in Tokyo confirm that the service has improved, though with other cellular services booming, it may be too late to rescue PHS.

Back in San Jose, Cooper and his colleagues are shifting their focus to a market that could be much larger: the wireless Internet. The company has designed a data-only version of its technology, called "i-burst." Now, working with makers of handheld computers, MP3 players, and other devices, ArrayComm wants to take palm-based computing into the broadband era.

Cooper says a leap like that is sorely needed. Palm VII devices from Palm Inc., for example, offer a modicum of Internet access. But they can only cruise a limited number of specially prepared Web sites, and they surf at a plodding 9.6 kilobits per second. Other companies in the U.S., including cellular services, are starting to offer the same kinds of access. "But this is the equivalent of taking a wide-screen, panoramic, stereophonic movie, and watching it through the peephole of your apartment door," says Cooper. ArrayComm antennas, widely deployed, could quickly lift that speed to one megabit per second, Cooper believes.

That kind of speed will never be delivered by 3G phones to a large number of subscribers, Cooper says. And many other engineers are inclined to agree. "You won't be watching 90-minute movies on your 3G phone," says Mark F. Bregman, general manager of IBM's "pervasive computing" business, which is in charge of the company's thrust into wireless. But with smart antennas, a network of handheld computers might be able to deliver such services, according to Cooper. At the very least, "you'll be able to download a five-minute song in about 20 seconds, wherever you are," he says. "If you could do that, why would you buy another CD?"

Funny he should mention music. In April, ArrayComm won a big vote of confidence from one of the world's most popular vendors of gadgets, games, and music: Sony Corp. Last year, Sony backed out of the U.S. cell-phone market, having failed to win significant market share. Now, Sony has taken an $8 million stake in ArrayComm, and Sony's new president, Kunitake Ando, has identified wireless-Web access as a top companywide priority. This week, the FCC gave ArrayComm a license to start commercial trials in San Diego.

ArrayComm has had no trouble attracting top-tier engineers. Four months ago, it lured a prominent radio engineer at Lucent's Bell Labs, Nitin Shah, to serve as ArrayComm's executive vice-president and general manager of Internet products. And soon, the company will announce a new raft of partnerships.

Just like any other startup, ArrayComm is facing all the usual investment and marketing hurdles. But with Cooper at the helm, it will be drawing on a long legacy of experience and credibility.