Unlicensed transmitters on TV broadcast frequencies
Back in 2004, the FCC proposed to allow unlicensed low-power transmitters on unused TV frequencies. Computer companies like Intel and Cisco thought it would improve computer networking and was the greatest thing since sliced bread. Broadcasters thought it would create interference and was a terrible idea. The cable industry had problems with possible interference to off-air reception of distant TV signals. Now the FCC has released a schedule calling for the start of deployment of unlicensed devices in February 2009, coincident with the end of the digital TV transition. So the interference issues must be resolved in a little over two years–less than that, actually. I'm not holding my breath.
Jeffrey Krauss 
But "listen before talk" is practically useless for protecting TV reception from interference. It may work to protect two-way communications or radar systems from interference when the communications or radar transmitter and receiver are at the same location. But TV receivers (including those at cable TV headends that receive distant off-air TV signals) are not transmitters, and unlicensed devices can never detect receivers. Moreover, an unlicensed device with an omni- directional receiving antenna might fail to detect a weak off-air TV signal that is providing adequate signal strength to a consumer TV receiver with a directional antenna.
And then there is the adjacent channel interference issue. The unlicensed device might detect a TV station on channel 19, but decide there is no TV station using channel 18 or 20, and therefore decide to use frequencies that fall into TV channel 18 or 20. That could cause interference in a TV set that is tuned to channel 19, because of the way TV set filters are designed.
So interest in "listen before talk" waned, and the FCC then proposed that devices should incorporate a geo-location capability like GPS and should access a database to identify vacant channels at their location. They could, for example, receive a control signal carried by an FM or TV station that downloads the frequency usage database to the unlicensed devices. But who would administer the database? Would the FM or TV station willingly give up channel capacity for free?
Another problem–the TV channels are widely used by unlicensed wireless microphones. As a practical matter, to avoid receiving interference, these new devices will have to avoid frequencies the wireless mikes use. But unlike TV stations, there is no FCC database of wireless mike transmitter locations.
As a sign of activity, the FCC has released a schedule. First, more industry comments will be solicited on a variety of unresolved technical concerns. Next, the FCC will complete measurements of the interference rejection capabilities of digital TV receivers, and will publish results of the interference studies it has underway.
The FCC claims it will publish the final technical rules for these unlicensed devices about a year from now. No way.
Meanwhile, the IEEE 802.22 committee is working on a standard for these products, which they call Wireless Regional Area Networks. Maybe the IEEE standard will be consistent with the FCC's technical rules...Or maybe not. The IEEE work deals not only with avoiding interference to broadcast reception, but also with designing a standard that allows the unlicensed devices to communicate reliably and securely with one another, and avoids interference to other unlicensed devices, at product prices consistent with other consumer electronics products. Not a trivial problem.
The FCC expects that manufacturers will start submitting products for FCC Equipment Authorization starting in December 2007, a year from now. That's wildly optimistic.
My feeling is that if these products are designed properly, so that they avoid causing interference, and if the FCC actually enforces its technical rules, then there is a pretty good chance that they can be used in rural areas where there are few off-air TV signals and few wireless microphones. But in the major population areas, after taking into account both co-channel and adjacent channel interference protection needs, there won't be enough free spectrum to satisfy potential demand, leading to consumer confusion and dissatisfaction. And the rural market alone won't be big enough to support prices in the range of consumer electronics products.
The FCC first adopted spread spectrum rules in 1985, but it took the WiFi folks several tries and many years before they finally got it right with 802.11g. So we'll see how long it really takes with this product. As I said, I'm not holding my breath.