Microsoft, Harvard researchers propose ‘WhiteFi’ network for UHF white spaces

Researchers from Microsoft and Harvard University have presented a detailed design and implementation of a wireless data network for use in unused portions of the UHF spectrum, commonly referred to as white spaces.

Laying out their proposal in a paper presented at the SIGCOMM 2009 Aug. 17-21 in Barcelona, Spain, the authors acknowledge the requirement the FCC placed upon wireless white space devices not to interfere with incumbent spectrum users, including TV stations and wireless mics, when authorizing their use in November 2008.

In “White Space Networking with Wi-Fi like Connectivity,” the authors say they have identified the challenges of building a UHF white space wireless network, dubbed a “WhiteFi” network, and have presented ways to surmount those challenges, including new techniques, algorithms and protocols “backed up by extensive evaluation over a prototype network” and simulations.

Throughout the past couple of years, the broadcast industry, wireless mic manufacturers and others fought and lost a battle with the FCC to keep roaming unlicensed devices out of unused TV spectrum. The FCC began evaluating prototypes in 2007 to determine if they could effectively detect and avoid spectrum in use by broadcasters, wireless mics and others. It called off its initial test midstream after submitted devices generally failed to do so. Following a second round of FCC tests as well as field tests at a sports stadium and in the Broadway theater district of New York City, the commission authorized use of the devices in TV white spaces as long as they did not create harmful interference to TV transmissions, wireless mics and other incumbent users of the spectrum.

According to the paper, the authors have built a WhiteFi network that can adapt itself to operate in the most efficient white spaces available while avoiding creating harmful interference. Major components of the network include a new spectrum assignment algorithm to manage variable bandwidth communications; a new discovery mechanism to be used by network access points; and a new way to handle unexpected disconnections.

Central to the researchers’ work is a new algorithm they designed called Signal Interpretation before Fourier Transform, or SIFT. According to the researchers, the algorithm is used in two ways: to analyze incoming signals in the time domain to find transmissions over different channels without requiring a wireless card to change its channel width and to reduce how long it takes to find WhiteFi access points that have switched to a different portion of spectrum. The researchers also have developed a new technique to let WhiteFi clients signal WhiteFi access point disconnection without creating harmful interference to ongoing wireless mic transmissions, the paper said.