Airwaves Up for Grabs: How much free space is left in the broadcast spectrum?

From Chris Wilson in Slate:

Just a day after Verizon Wireless spent nearly $10 billion in its bid for a valuable slice of the airwaves last week, Google asked the Federal Communications Commission to open up other unused pieces of the spectrum for wireless broadband. The plan calls for allowing companies like Google and Microsoft to beam wireless Internet access on frequencies between those allocated for television channels—in the so-called "white space"—as well as frequencies reserved for channels that don't exist in a given area. How much of the broadcast spectrum is still up for grabs?

It depends where you are. The "broadcast spectrum" refers to a portion of the full electromagnetic spectrum that is ideal for telecommunication, with frequencies much lower than infrared or visible light. Federal law grants the FCC the authority to determine who can broadcast on which frequencies between 9 kHz and 400 GHz, i.e. the entire range of radio waves and microwaves, to prevent interference between stations. For example, the 410 MHz band is reserved for radio astronomy, while the range from 88 to 108 MHz is for FM radio. (If the government didn't keep track of who broadcast in which frequencies, there would be tremendous interference between broadcasts, making a clear signal very difficult to find in congested areas.) But frequencies allocated by the FCC aren't always in use. Whether a given region of the spectrum is occupied depends on the size and demand of the local population. An urban area with a lot of broadcast stations might fill up most of the spectrum allocated for radio and television, while a rural area would leave much of it unused.

Google's white-space plan concerns television broadcast frequencies, which are divided up by channel throughout the spectrum. The chunks that the FCC just auctioned off to Verizon and others, in the 700 to 800 MHz range, have long been reserved for television stations broadcasting analog signals. But once TV broadcasting goes fully digital in February 2009, the stations will clear out of those frequencies. Meanwhile, companies are interested in using parts of the spectrum that are already allocated, but not always occupied. To accomplish this, they'd need to produce devices that can search for competing signals and suss out any frequencies that happen to be vacant. Proponents like Google say the vast majority of the airwaves go unused most of the time and will remain so until these devices are widespread.

So far, early testing of these "White Space Prototype Devices" has not gone particularly well. In an initial round conducted in July 2007, two prototypes were either unable to detect competing signals or detected signals that were not actually present. (Microsoft claims they sent a defective version of their model to the FCC.) This poses a real problem for the white-space plan: If a device tries to initiate a broadcast at the same frequency as an existing signal that it failed to detect, it could cause interference. Digital broadcasts might begin to skip or freeze, like a scratched DVD. Opponents of the white-space plan, including the National Association of Broadcasters, cite these reports as evidence that the technology is not ready for public consumption. The FCC is currently conducting a second round of tests.