Dr. David Reed: More to the radio spectrum than meets the ear

Dr. David Reed, inventor of TCP/IP and the "end-to-end" theory, is one of the cornerstone theorists behind the construction of the Internet. Today, he's twisted his dial over to another medium: He wants to do for the radio spectrum what the Internet did for the network. At the very least, he wants it opened wide for reexamination. And if he has his way, new regulation for the 21st century will indeed happen—but probably later than sooner. Government doesn't change direction on a dime, unlike the networked world, and there are powerful forces arrayed against change as well.

Reed addressed long-entrenched FCC restrictions on radio, which has yet to develop any kind of meaningful relationship with the Internet, at the recent O'Reilly Emerging Technologies conference in Santa Clara, CA. Essentially, Reed wants to open-source at least part of the radio spectrum and protect it through the new Creative Commons organization.

"In radio, waves pass through one another without damaging each other. Interference is a radio being confused and not being able to decode the signals," Reed says. "The FCC regulates (radio) as if the capacity is equal to the bandwidth." Reed points out that the FCC was created on the belief that there's not enough bandwidth to accommodate the constitutional guarantee of free speech in broadcast communications. He says technology changes all that.

Stretching the radio spectrum
Reed and his colleagues theorize that the radio spectrum is like an accordion: It stretches as needed to play the notes you want to play. They believe technology has long left the old AM/FM radio world in the dust.

"The key idea is to organize the spectrum users into a cooperative network rather than an uncoordinated set of point-to-point channels," Reed writes on his Web site. "As the number of users N increases, the capacity can grow as N1/2 (the square root of N) in the architectures they suggest. Thus, the more users that make up the network, the more capacity the network can carry."

The time is now for the radio industry and the FCC to move into new-century thinking and look at the spectrum with unbiased eyes.

"What we haven't done is gone back and reexamined if technology can fix some of the problems created by handing out bits of spectrum to special interest groups," Reed said at the conference. "You do this by collaboration among radios/devices—essentially, software-defined radio. Is there potential for a problem (enough signal to burn out all others)? Yes, some regulation is necessary, but not the kind we have today."

Software-defined radio reinvents broadcasting
Most of that regulation would be in the power of broadcast megahertz. And what is software-defined radio (SDR)? "You write software and run it on devices (presumably handheld or desktop) that help users find the bandwidth they need to do what they want to do: broadcast their words, music, anything," Reed told the conference audience. "You can publish what you want to the world via the Web using Web pages right now; why not be able to use radio for the same purpose?"

Afterward, Reed spoke to me in greater detail about how software-defined radio would work.

"Software-defined radio is an evolving set of techniques that allow a device to emit and receive electromagnetic waveforms (i.e., radio waves) under full software control, with very little limitation on the waveforms it can use and understand," he said. "Instead of building a radio where the hardware only works with certain kinds of wave shapes, frequencies, and so forth, a software-defined radio can vary its behavior dynamically in a short time period. So a single SDR can behave like any existing hardware radio (by loading the right program), and like kinds of radios that we have never seen before, just by changing the software."

Multiple SDRs can form networks, he said, that dynamically adapt the waveforms they use to fit their environment and the traffic demands.

Plenty of bandwidth for everyone
Reed insists there's plenty of bandwidth available for everybody's needs. To explain this in detail, he's added an Open Spectrum page to his Web site.

"We need to do for spectrum what the Internet did for the network," says Reed. "If technology improves, the regulations become unnecessarily strict. When we make better radio systems, the old ones become obsolescent and wasteful. We could improve the entire system by junking the old stuff, and replacing it all with functionally compatible systems that are based on new insights and design."

Perhaps, but don't hold your breath. Current owners of conventional radio bandwidth—the rich major television networks not the least among them—pay big bucks to Uncle Sam for their positions on the dial each year and aren't going to let others, even scientists, elbow in on their "territory," as they see it, for little or nothing.

However, Reed and other visionaries believe there's more than enough radio space to go around. The theory, he says, is that bandwidth increases along with the number of receiving units, just like the Internet.

This could be the start of something big
"Essentially, all the (conventional) spectrum is now owned. What I would like to see is a part of any additional spectrum we find be made open, for anybody to use," he said. "The FCC's mission is to see that the radio spectrum is utilized for the 'public good,' but it's not clear that it's all being used for the public good right now. We need some spectrum in which we can do anything we want."

Cheers erupted from the standing-room-only O'Reilly conference audience.

"Scalable wireless networks change the rules from scarcity to cooperative creation of abundance," Reed said. "It feels like there's a movement beginning, and this could be as big as the Internet all over again, this time enabling mobility and reconfigurability."

There may indeed be the beginnings of a movement. We know a lot more about scalable networks now than we did just five years ago. But don't count on anything happening soon: There are too many powerful, entrenched businesses—and a very slow-moving FCC—that are quite happy living in the 20th century.