NEWS RELEASE, 1/28/99
New Berkeley Wireless Research Center aims to
put a radio in every office computer and home appliance
By Robert Sanders, Public Affairs
Berkeley -- To most people, "wireless" means scratchy cellphone connections that cut out far too often, or the occasional laptop radio modem that provides an agonizingly slow internet link.
A new center opening this week at the University of California, Berkeley, is setting its sights on improving wireless for the generations to come. Their goal is to develop the technology for next century's wireless revolution, which will allow complete radio systems to be put on a single computer chip.
The center's research goals include creating universal radios that efficiently share scarce space in the radio frequency spectrum; designing tiny, cheap "picoradios" that can be embedded into every household appliance, adapt to any network and have a battery life of years; and developing ultra-high frequency radios that will allow inexpensive use of now unused regions of the frequency spectrum.
Wireless has been growing by leaps and bounds as mobile phone networks expand around the globe. A recent industry analysis estimated that by the end of 1998 nearly one quarter of the U.S. population would have wireless phone service, and that in 1999 some 80 million Americans would be signed up.
This is only the beginning of the wireless revolution, however. Its main impact will be felt when data is sent over the air and internet access will be available any place, any time.
The Berkeley Wireless Research Center is supported by six corporations that today lead the world in wireless communications, and by several million dollars in government research funding from the Defense Advanced Research Projects Agency (DARPA). Each of the industrial members has contributed $450,000 over three years to create the center at Shattuck Avenue and Allston Way in Berkeley.
Occupying an entire second-floor suite, the center will house between 40 and 60 researchers from UC Berkeley, other universities and from industry, including graduate and undergraduate students and seven faculty from UC Berkeley's Department of Electrical Engineering and Computer Sciences.
A companion effort, the Gigascale Research Center, funded by the Semiconductor Industry Association and DARPA, also will be partly housed at the center. It focuses on future design tools to cope with circuits that contain hundreds of millions of transistors.
The center is to be dedicated Friday, Jan. 29, by UC Berkeley's Associate Vice Chancellor for Information Systems & Technology Jack McCredie, the College of Engineering, faculty and student researchers and the supporting companies. The industry members span all the critical areas needed to carry out the research agenda: Cadence Design Systems, Inc., a developer of software to design computer chips; Intel Corp., Texas Instruments, Inc., and STMicroelectronics- semiconductor companies that have the world's most advanced chip fabrication technologies; and Ericsson, Lucent Technologies and Hewlett-Packard Co. - multi-billion dollar companies that provide both telecommunications systems and components.
These and other companies have also donated equipment to the center to make it the premier facility for doing wireless research. The donations include high performance computers and servers from Intel; an interior cellular phone system and high speed network infrastructure from Nortel; test equipment from Hewlett-Packard; software from Cadence; a large RAID (redundant array of inexpensive disks) file server from Network Appliances; and an indoor wireless GSM base station from Ericsson.
"The new wireless center represents a unique approach to create a better, more effective partnership with industry," said College of Engineering Dean Paul Gray, a participating faculty member. "Having university faculty and graduate students working closely with industry people has tremendous potential for advancing the basic techniques needed to implement wireless communications systems."
"The center will concentrate on areas far ahead of present corporate developments, which is one of the reasons that it is necessary to have such tight industry interaction," said Gary Kelson, technical director of the center.
"The real future of wireless is in data," said Robert Brodersen, professor of electrical engineering and computer sciences at UC Berkeley and co-scientific director of the center. "Providing widespread data access for applications that are unknown today requires a much more flexible approach to using the air waves. The present approach involves setting standards which are essentially obsolete by the time they are finalized."
To solve this problem, a major effort in the center will be to develop a "universal radio" that would require only a loose "etiquette" - the hand-shaking or protocol required to establish a network connection. This would allow many users to simultaneously use the same frequency bands without interference.
The goal is to change the way the Federal Communications Commission sets the standards for frequency bands. To assist in this ambitious goal the center has teamed up with researchers from the Wireless Network Research Lab (WinLab) at Rutgers University and theory researchers at Princeton University. Co-director Jan Rabaey, a professor of electrical engineering and computer sciences at UC Berkeley, focuses on design methodologies that will allow rapid and low-risk design of the sophisticated "wireless systems on a chip", so that actual radio prototypes can be designed and tested at the center and then fabricated by industry sponsors.
A second emphasis in the center will be development of picoradios, which will be on a single chip and consume 10 to 100 times less power than today's radios.
"Picoradios will allow wireless links to be embedded into all sorts of small devices and appliances, making possible truly 'smart homes,'" Rabaey said. "Since there will be so many of these devices they will need to be very inexpensive, last for years without a battery replacement and automatically self-configure, so that the user will not be aware of the network that has been set up."
A third project aims at investigating what new frequencies can be used as integrated circuit or IC technology improves over the next ten years. At higher frequencies, in the 50-60 GHz (gigaHertz) range, there is much unused spectrum that will eventually be needed when use of broadband wireless data links becomes widespread.
"Unfortunately ultra-high frequencies can't be used cost effectively now because radios operating at these frequencies require exotic materials which are very expensive," Kelson said. "We will investigate the use of standard CMOS (complementary metal oxide semiconductor) technology to see when it will be possible to inexpensively exploit these higher frequencies."
Many of the UC Berkeley faculty involved in the center were part of a six-year project, called InfoPad, whose goal was to develop the hardware, software and mobile network support which would allow wireless access to high speed networks and the internet from inexpensive, portable terminals. That project, supported by many of the same sponsors as the Berkeley Wireless Research Center, ended a year ago.
Two of the primary researchers on that project, Brodersen and Rabaey, approached corporate sponsors about a more elaborate research project that would extend the InfoPad work on wireless technology. They proposed to pull in other research under way at UC Berkeley, ranging from chip design and computer aided design (CAD) tools, to the design of wireless network architectures and applications.
Wireless companies were keenly interested.
"The sponsors are all world-class companies, and we want this to
be a world-class facility with research which fundamentally changes the
nature of wireless communications," Kelson said.
Send comments to: email@example.com