Friday, March 28, 2008

Google's Wi-Fi Dreams

Tech giants and broadcasters tussle over vacant TV channels.

By John Borland

The new Wi-Fi: Next year's digital-TV transition will free up empty TV channels for use by broadband companies. Google wants to use them to create what it calls Wi-Fi 2.0 services. Credit: Technology Review

With next year's transition to digital television sparking fierce debate over how to use the newly vacant TV channels, Google has offered a plan that it says could vastly improve U.S. broadband service.

For several years, policy makers, technology companies, and broadcasters have been debating the best use for this newly available wireless real estate. This "white space" between operating TV channels is of particular interest to broadband companies, because wireless signals sent at these frequencies will have the ability to penetrate walls and other obstructions more easily than do cell-phone or Wi-Fi signals.

Google submitted its new proposal to federal regulators late last week, outlining a plan to utilize the unused channels for what the company calls Wi-Fi 2.0: a loosely regulated set of broadband services with the potential for gigabit data speeds. The company stopped short of saying that it would seek to operate a broadband network itself, but it clearly sees business potential, telling federal regulators that it would provide other companies with "the technical support necessary" to turn the TV frequencies into broadband data conduits, free of charge.

"The unique qualities of the TV white space--unused spectrum, large amounts of bandwidth, and excellent propagation characteristics--offer a once-in-a-lifetime opportunity to provide ubiquitous wireless broadband access," wrote Google attorney Richard Whitt in the company's proposal to the Federal Communications Commission (FCC).

Indeed, the soon-to-be-freed TV airwaves represent some of the last, and potentially most valuable, swathes of U.S. wireless spectrum still suitable for providing broadband services. Naturally, the debate over how they will be used--and who should use them--has been fierce.

Technology companies such as Google, Microsoft, and Motorola say that these frequencies are particularly well suited for providing rural areas with high-speed Internet service, as well as for short- or medium-range networking applications that might provide data transfer rates of gigabits per second, as opposed to the roughly 54 megabits per second of today's 802.11g-based Wi-Fi networks.

Regulators have already approved the use of some fixed devices, but the tech companies want approval for mobile devices such as handhelds and laptops as well, and they're seeking rules that would let companies offer services using the spectrum without having to get new licenses. Google has said that the TV white space would be ideal for mobile data devices using its open-source Android platform, for example.

However, TV broadcasters are worried that unlicensed devices sending data on unused TV channels--say, a vacant channel 29--might interfere with the program signals being broadcast on channel 28 or 30. Similarly, they're concerned that mobile devices able to operate across a range of frequencies might accidently choose a channel being used locally for TV service and scramble nearby TV viewers' signals.

Technology companies say that they can control for this problem. They've advocated a technique called spectrum sensing, in which the portable devices, transmitters, and receivers alike would scan channels before using them to make sure that they are indeed vacant. Detecting the presence of TV signals or other authorized users would keep a "white-space device" from using that channel.

However, FCC tests of spectrum-sensing prototype devices have been spotty at best. A first round of testing last year had disappointing results. Another round of FCC testing, with a new batch of prototype devices from four different companies, is now under way.

Google's new proposal aims to deflect concerns about those tests. To this end, the company has adopted three alternative ideas that were previously advocated by Motorola and others for protecting against interference.

The first aspect of this protection would create a publicly accessible database listing all licensed TV stations and their geographic location. Any device attempting to use the TV spectrum would first have to establish its own geographic location, by using GPS readings or another means, and then check this database to avoid conflict with a licensed TV station in that area.