Policy —

White spaces on steroids and free spectrum for everyone

Unlicensed spectrum brought us wireless phones and WiFi—so why isn't more …

Say you have some bright idea for the "next WiFi" and you just need a tiny little smidgen of open spectrum in which to deploy the invention that will bring cheap, easy, ubiquitous communications nirvana to everyone. Can you get it?

Generally, no. The US government squats on huge swaths of spectrum, while paid-up license holders (like cell phone service providers) control much of the rest. Slivers of spectrum are left open for unlicensed use, and those tiny bands have produced great big social benefits: wireless baby monitors, wireless phones, and WiFi.

But a set of papers from the New America Foundation argue that the Obama administration should take a different course on spectrum, making it simple for entrepreneurs to launch new wireless devices even in occupied bands. Their common credo: spectrum is abundant—if you treat it right.

Let's start sharing.

The virtue of sharing

The problem, as all three papers see it, is the old command-and-control model of passing out licensed spectrum and giving exclusive control of the band to license holders. This has resulted in spectrum going to federal entities like the Supreme Court, for instance, which has been assigned 13 bits of spectrum; the National Gallery of Art has 12, while the Smithsonian has 73.

And these are bit players. The Air Force has 28,227 bits of spectrum, while the overall Department of Defense controls the single biggest chunk of spectrum in the country.

These sorts of carve-outs have created a spectrum map of almost unimaginable complexity (see below for the most recent one from the government).

spectrum_map.png

Most of these grants are national, even if the organization in question only uses them in a few miles of space. They are also exclusive, even if the spectrum user broadcasts only once or week, or only uses specific angles and directions, etc.

All three papers want to move spectrum policy to a "commons" model in which the complete spectrum is mapped to find out how and when each band is actually being used; unlicensed users could then use many of these bands when they are unoccupied.

Here's how specific this sort of sharing could be. Big bands of spectrum are reserved for radar use (weather, airport control, etc.), but radar generally broadcasts from a spinning tower. Most of the time, then, a radar installation is facing away from any given location. Using a smart radio with knowledge of the transmitter's behavior, wireless devices could share the band with the radar systems, shutting down their own transmissions at the precise moment the radar swings round to face them, then resuming as it moves on.

As New America's Michael Calabrese puts it, "the vast majority of frequency bands are not being used in most locations and at most times."

Smart radios, comprehensive databases

The solution is smart radios that can operate in many different bands, listening for existing transmissions and going quiet when they are found. The radios might also tap into a vast database that tells them what spectrum to use or avoid in specific locations of the country; if the Air Force uses a frequency for missile testing, but only in three US locations, everyone else could make use of the band thanks to the database.

If this sounds a whole lot like the "TV white space devices" that were such a contentious issue before the FCC last year, they should—the idea is exactly the same.

In fact, the new papers could basically be summed up as "white space devices on steroids." That is, the same technology would be deployed, but it wouldn't be limited just to TV bands—devices could use huge chunks of lightly used spectrum thanks to spectrum-sensing tech and a beefed-up white spaces geolocation database.

Calabrese points to a 2004 test showing that even in New York City, most spectrum lies fallow most of the time. "The highest occupancy on the prime beachfront spectrum below 3GHz was just 13 percent in New York City," he says, "while the average across locations studied was just six percent."

Making that spectrum available to other users—possibly even on a real-time micropayment model that reimburses the license holder—would go some way toward creating truly "pervasive connectivity," says Calabrese.

It's a compelling vision of the future, but interference concerns are real and license holders will often be loath to share spectrum. But some won't mind. In fact, the Department of Defense already allows such sharing on certain bands that it controls but uses infrequently.

New America scholars were big backers of the successful white spaces initiative at the FCC and are now pushing their ideas further out into the spectrum map. With Google's Eric Schmidt running the think tank and a host of Obama staffers with similar spectrum and connectivity goals, the approach is certainly one that could find favor with the government.

We'll have a much better sense of how well the geolocation database and spectrum-sensing technologies work on national scale once white space devices go on sale at the end of this year.

Further reading:

Channel Ars Technica