Wireless at WARP speed

Nothing kills innovation like reinventing the wheel. Until several years ago, electronics researchers testing new high-speed wireless technologies had to build every test system from scratch.

“It was incredibly frustrating,” said Ashutosh Sabharwal, director of Rice University’s Center for Multimedia Communication (CMC). CMC set out to change that in 2006 by creating a turnkey, open-source platform.

In two years, the platform—dubbed WARP—has attracted the attention of Nokia, MIT, Toyota, NASA and Ericsson, and already it has been used to test everything from low-cost wireless Internet in rural India to “unwired” spacecraft.

Sabharwal, the lead investigator on the federally funded WARP project, said he and his CMC colleagues were among the lucky few in academia who could afford the high cost of entry into wireless research in 2006.

“Collectively, it was a big waste of time and effort, and there were a lot of people who simply couldn’t afford to play,” Sabharwal said. “Some of our previous research hinted at the possibilities of an open-access platform, so we had a clear goal when we made our proposal to the National Science Foundation.”

WARP stands for “wireless open-access research platform,” and it resembles the guts of a desktop computer. What makes WARP boards so effective is their flexibility. When researchers need to test several kinds of radio transmitters, wireless routers and network access points, they need only to write a program that permits the WARP board to become each device.

Motorola is using the system to test a new architecture for wireless Internet in rural India, and NASA is using WARP to look for ways to save weight, cost and complexity in the wiring systems in spacecraft.

Making WARP a reality wasn’t easy. Students and staffers from the research groups of Sabharwal and CMC faculty members Ed Knightly, Lin Zhong, Joseph Cavallaro and Behnaam Aazhang designed the WARP hardware and built the back-end systems, tools and software that allow various components of WARP to work together.

CMC was able get a version of WARP ready to release to the research community within a year of its initial NSF funding. After early success, Sabharwal spent months seeking a company to manufacture WARP boards.

“Our philosophy from the beginning had been to drive the cost lower and lower, to sell the boards for as little as possible in order to get them out there,” Sabharwal said. “Everyone we contacted seemed to want just the opposite, to mark them up as much as possible and sell to the few people that could afford high prices.”

With CMC researchers touting their work at conferences and workshops, colleagues around the world expressed interest in the boards. Sabharwal said CMC began producing a few, even as it was seeking a production deal with an established company. The lab wound up selling equipment to some 40 university and corporate research groups before a WARP architect founded Houston-based Mango Communications in mid-2008 to take over production of the boards.

Sabharwal said CMC has NSF funding through 2010 to further develop WARP, and will put the final touches on a new set of tools that will allow researchers to control the boards from any location remotely. That will permit them to fulfill one of CMC’s longstanding goals—installing the flexible boards into existing test networks like the CMC-built high-speed network that nonprofit Technology For All operates for more than 4,000 users in Houston’s East End neighborhood.

Jade Boyd, Rice News