David P. Reed
Position: Adjunct professor of media arts and sciences, MIT; HP Fellow, Hewlett-Packard Labs
Personal Point of Impact: Leading researcher in next-generation radio networks; public advocate of changing spectrum regulations to allow experimentation with new radio technologies Technology Review: Why is our use of radio causing problems?
David Reed: There’s clearly a huge demand for wireless digital communications that is driving high growth rates of services and devices, from traditional cell phones to Wi-Fi devices to other new things. We’ve gone from the idea where radio is an expensive thing that you only want to use when absolutely necessary to the idea that it’s a convenience item for interconnecting everything, so my mouse and my keyboard talk to my computer by radio. The flip side of that is, what if we start doing that more and more as it gets cheaper and cheaper? Does everything start interfering with everything else, and do we have to pick what’s allowed to talk to what? That’s the question: how do you meet this overwhelming demand and overwhelming possibility with a sensible way of scaling up the use of radio.
TR: “Radio” brings to mind what we turn on when we get into the car, where a station broadcasts at a certain frequency, and if someone else uses that frequency, then we can’t get our music or talk show. Is that not how all radio technologies work?
Reed: Well, it’s certainly not correct from the point of view of the technology. Long ago, when radio spectrum was wide open and we weren’t able to do very good radios, we decided that the best and cheapest way to allow many radios to operate on the same channel was to divide the spectrum up according to the application. So we have bands that are assigned to broadcast AM radio, bands for television, for two-way communication, and all that. We didn’t think at all technologically about that; dividing by frequency was easy to do given the technology of the day.
In the past 10 to 12 years, we’ve started to realize many, many technologies can effectively share the airwaves without necessarily causing each other to malfunction. But the regulations that we and other countries apply to radio transmission don’t admit that those new approaches are even legitimate. To get a new technology approved, especially one that contradicts the original assumptions, is virtually impossible-an incredibly political problem with lots of vested interests in keeping things the same.
TR: How does that play out practically with the Federal Communications Commission, which regulates airwaves in the United States?
Reed: A really interesting example is what happened to ultrawideband. The company I was at before 1996-Interval Research-was looking at innovative technologies for very-short-distance networking, and one of the things we discovered, in its infancy, was ultrawideband. Interval financed a large investment in taking that technology from a research possibility to some commercial capabilities and spun off a company called Fantasma Networks.
Ultrawideband is a technology that uses very, very low energies in every band, so it theoretically should not interfere with existing services. But proving a negative-proving that it will not interfere with any existing service-is extremely difficult, and the FCC basically had to push ultrawideband through the approval process without that absolute certainty. For several years, the FCC delayed rule-making on ultrawideband. The prospects of interference were the big issue. But underneath it all, there’s also this issue of, if it was made legal, many services could move to ultrawideband, and then there would be unrestricted competition against existing providers of services. For example, you might see ultrawideband being used to provide television, or radio, or two-way telephone communication, and that would eliminate the monopoly or oligopoly benefits that come to the current spectrum holders. So they have a strong incentive to fight this. It’s very hard to make a wise decision confronted with highly politicized technical arguments.
Fantasma continued to develop the technology, but approximately a year before the FCC finally allowed some ultrawideband, the investment community lost patience in this company, and it basically had to fold. It shows how the regulatory environment really stifles potential innovation that is crucial to solving this huge need of having much more capacious wireless networking capabilities that can work at all distances and scale to much larger numbers of devices and users than we have today.
TR: Are there other disadvantages to the current regulatory structure?
Reed: There are several problems. One problem is that it’s very hard to determine in advance what services will be successful. So if you try to put the FCC or some international body in charge of determining which are the most beneficial new technologies, you’re putting the cart before the horse. If you can’t try it out on real customers, then you have no idea what’s going to work.
The second thing about the regulatory system is that-most people will agree-it’s been captured by those that it regulates. If the FCC proposes any new thing, the first [group] they hear from is the lobbyists, who use a variety of arguments to either say it threatens their businesses or it threatens their technology. In the case of radio, it’s almost always formulated as an argument about interference, even if the real issue is competition. And it’s very hard to refute certain technical arguments, especially in a forum like the FCC: the FCC does not have a strong independent technical evaluation capability.
TR: So do new technologies such as ultrawideband really not interfere with, say, my cell phone or an airplane’s electronics?
Reed: Left to ourselves, I think engineers would be able to find solutions to almost any of these problems-should they turn out to be problems. For example, there is a whole lot of unused capacity in the current UHF television band. There’s an FCC proposal to allow unlicensed use of that spectrum as long as the technologies being deployed don’t interfere with existing users. Because that band propagates much better through trees and other sorts of things, it could enable a new generation of digital communications networks that could coexist with television, and that would handle longer-range communications better than Wi-Fi. There are a lot of potential engineering solutions to avoiding any interference that might be caused. Software-defined radios are now possible that can essentially dance around the existing television services. But it’s very hard to make the case that this will work when you’re basically facing political opposition, rather than technical opposition that wants to work with you.
TR: So what could improve the situation?
Reed: I think the desirable future would create more opportunities for experimentation and follow-on. Right now there is a huge disincentive to even do research on novel kinds of systems, because it’s considered extremely expensive and almost a waste of time to try to go through the FCC process for anything that’s very different from what’s currently out there.
The other side of that is creating regulatory openings that allow significant exploration. A really good thing that happened was the opening of the completely unlicensed spectrum that Wi-Fi and the new 802.11a wireless standards work on; [the Wi-Fi band] was authorized by the FCC in 1985, and the [802.11a] bands were authorized in the mid-1990s. The technology gradually evolved, and a networking standard was developed to increase interoperability in the late 1990s. We need more of these unlicensed bands-and of wider frequency-because people want more and more devices to communicate faster and faster, and today’s Wi-Fi architecture doesn’t scale to handle more users efficiently. New bands increase the rate at which innovations can be brought to market.
The big new idea for making self-driving cars that can go anywhere
The mainstream approach to driverless cars is slow and difficult. These startups think going all-in on AI will get there faster.
Inside Charm Industrial’s big bet on corn stalks for carbon removal
The startup used plant matter and bio-oil to sequester thousands of tons of carbon. The question now is how reliable, scalable, and economical this approach will prove.
The hype around DeepMind’s new AI model misses what’s actually cool about it
Some worry that the chatter about these tools is doing the whole field a disservice.
The dark secret behind those cute AI-generated animal images
Google Brain has revealed its own image-making AI, called Imagen. But don't expect to see anything that isn't wholesome.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.