Some of the challenges that faced the group came about because of the undefined nature of white-space frequencies. . The researchers designed their algorithms to determine the ideal amount of frequency bandwidth to use for a broadcast, balancing the desire for strong signal against the possibility of interference with neighboring frequencies. They also had to design a way for mobile devices to find a signal from an access point.
One of the most important parts of the White Fi system is a protocol for dealing with collisions among different signals (particularly those from wireless microphones, which can turn on at any time). Even a single packet of interference is enough to produce audible disruptions for a microphone. Even if interference affects only one device on the network, strict regulations forbid all devices on the network from using that channel. The researchers got around this by designing the access point so that it maintains a backup channel. If another user is detected, the white-space device or access point immediately switches to the backup channel, which reassigns bandwidth use as needed.
Peter Steenkiste, a professor of computer science at Carnegie Mellon University who specializes in networking, says that previous work on white spaces has focused on addressing one problem at a time. “The thing that I think is very interesting about this paper is that it really has looked at how you put a complete system together,” he says.
Steenkiste adds that “there are a lot of practical issues that they’ve worried about.” In particular, he says, the researchers did not assume an ideal, controlled environment for their system. Rather, they took into account such problems as measurement “noise” and the unpredictable behavior of wireless microphones. “[The research] has an answer for every question,” Steenkiste says.
Chandra says that his group recently received an experimental license from the FCC that allows them to build a prototype White Fi system on the Microsoft Research Campus in Redmond, WA. They plan to send their findings to the FCC in the hope that the data will help determine future white-space regulations. Chandra notes that since the transition from analog to digital television is happening worldwide, there is a high level of international interest in US white-space experiments. Researchers and companies all over the world are looking for technologies to take advantage of the fragments of spectrum that will open up in the coming years, he says.