As more and more wireless gadgets suck up more high-speed Internet bandwidth, the U.S. Federal Communications Commission (FCC) is challenged to find new avenues for consumers to get their mobile connectivity. One controversial option that FCC Chairman Julius Genachowski is considering as part of his agency's National Broadband Plan is to make slots of "white spaces," unused airwaves on the broadcast spectrum, available to the unlicensed mobile phones and PCs that are typically used by the general public.

These white spaces originally existed between analog television frequencies but were freed up when broadcasters switched their TV signals to digital in June. Still, critics fear that a flood of unregulated wireless devices will interfere with one another or with licensed users, such as wireless microphones used during newscasts, Broadway performances or professional sports events.

If unlicensed white-space use is written into the final plan, the challenge will be developing technology that can accurately locate available airwaves while avoiding that interference. Unanswered questions over how this will work are one of the reasons Genachowski asked Congress last week for a one-month extension (pdf) (to March 17) for filing his plan.

There are essentially two proposed approaches to finding white spaces in the spectrum and determining whether they are already being used by other wireless devices or by license-holders. The first is for the FCC or some private entity to create a national database listing all owners and operators of wireless microphones and their RF (radio frequency) channel assignments. A cell phone or laptop would consult the database regularly, locate an open space in the spectrum, and calibrate to connect to the Internet via that space.

The U.S. House of Representatives is already pushing a bill that would require the FCC to create such a database. Google, a company that stands to gain a lot by opening the spectrum to wireless devices, has asked to help monitor the database, although many have criticized this as move as tantamount to "allowing a fox to watch the henhouse," reports.

There are a few disadvantages to this database-lookup approach. "There are practical issues about who will run the database, who will add information (about spectrum use and availability) to the database, and who will pay for the database," says Luke D'Arcy, head of cognitive radio technology at Cambridge Consultants, a technology development and consulting firm. Another potential problem is the paradox that a wireless device would need to connect to the Internet to check the database but would not have access to the Internet until it first checked the database, he adds.

The second approach is for wireless devices to include a cognitive radio. Such a component would allow wireless devices to check the spectrum directly for available frequencies and connect to them right away. Cambridge demonstrated some of this capability through its Whitespace TV prototype last week at the Consumer Electronics Show (CES) in Las Vegas. Although the company is not planning to release Whitespace TV as a product, the device showed how Cambridge's InCognito cognitive radio can change television-signal reception parameters and wirelessly stream them to multiple TV sets over open white-space frequencies.

Cambridge is one of several companies developing spectral-sensing technology that would find open frequencies for over which wireless devices can connect.  In 2008, Adaptrum, Inc., Microsoft, Motorola, Philips Electronics and Singapore's Institute for Infocomm Research all presented prototype white-space detection devices to the FCC for evaluation. These devices performed with varying levels of success.

"The problem (with spectral sensing) is you have to be able to pick up very faint signals," D'Arcy says. "In practice it's very difficult to make this work reliably."

More likely, the FCC will back wireless devices that can monitor the proposed database while also scanning for available frequencies. "The consensus within the industry is that the [database-] lookup approach will work and that sensing hasn't been proven to work reliably enough at this time," D'Arcy says. "My personal view is that sensing will work. It just needs more time to be further developed from a technical standpoint."