The End of Spectrum
By Kevin Werbach, Thu Oct 02 12:15:00 GMT 2003

There are two basic problems with the rules governing the radio spectrum: there is no such thing as spectrum, and we don't need any rules.


As we all know, wireless communication can occur because governments parcel out access to the radio spectrum, a limited resource divided up into frequency bands. Without governments partitioning spectrum, there would be a cacophony in which no one could communicate effectively. Agencies like the Federal Communications Commission in the US must decide who can do what, in which frequencies. The alternative is chaos.

Don't be so sure.

Everything I've just said is a lie. Familiar lies. Lies which were probably worth telling eighty years ago, when the basic structures of wireless regulation were established. But lies nonetheless. It's time to tell the truth about wireless communication, and to change the fundamental framework of wireless policy.

Spectrum is an intellectual construct that helps us grasp a deeply alien phenomenon. We take wireless communication for granted because it is such a significant element of our lives, and has been for a century. But how many of us really understand how it works? Radio signals are electromagnetic radiation, governed by the mysterious laws of quantum mechanics. Even Albert Einstein could only explain radio by describing what it is not:
"You see, wire telegraph is a kind of a very, very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? And radio operates exactly the same way: you send signals here, they receive them there. The only difference is that there is no cat."
Einstein's "cat" is spectrum. It is the physical medium we imagine carrying radio waves from one point to another. When governments assign exclusive licenses to wireless frequencies, they are carving up the cat among a limited number of users. (One must hope it has at least nine lives.) Yet there is no cat. Regulators are actually just telling owners of wireless devices at both ends of the transmission what they can do with their equipment.

Why does this distinction matter? Because, as deep thinkers such as Yochai Benkler and David Reed have argued, the spectrum metaphor no longer holds up the way it once did.

At the end of the 19th century, Guglielmo Marconi developed a technique for multiplexing radio signals based on frequencies. (Before that, only one radio could operate in any location.) Using frequency bands to separate systems was a design choice, not something inherent in wireless communication. For many years, though, frequency division was the only mechanism that was technically and commercially viable. If two radio or television transmitters were physically close together and broadcast on the same frequency, receivers would experience interference. Therefore, it was reasonable to regulate those services by giving them exclusive licenses for frequencies.

Wireless technology has come a long way since Marconi’s day. For example, many systems now use spread-spectrum as an alternative to frequency division. WiFi and CDMA cellular networks are two examples. In both cases, devices share the "same" frequencies, using computational intelligence to distinguish one signal from another.

The ultimate form of spread-spectrum is ultra-wideband (UWB). UWB systems can transmit at extremely low power, so low that they appear invisible to licensed systems operating in the same bands. The US Federal Communications Commission authorized commercial deployment of UWB for the first time in February 2002, and devices should start coming on the market next year. Other novel forms of wireless sharing include angle-of-arrival (for example, distinguishing a satellite transmission coming from overhead and a terrestrial wireless signal moving horizontally); mesh networking (relaying signals from other users); and cognitive radio (sensing the environment to find temporarily open "holes," and moving out of the way once another signal appears).

If frequency division isn't the only technical mechanism for differentiating wireless systems, why should it be the primary method of regulating those systems? With today’s technology, rules that grant exclusive control over frequencies make spectrum artificially scarce. They prohibit alternative approaches that would not actually create harmful interference.

The problem is that any set of pre-defined parameters will eliminate some systems that could increase wireless capacity. Whether a new wireless system can coexist with existing systems is a fundamentally contingent question. It depends on the technical characteristics of all the systems involved, and on the types of services those systems provide. There's no way to build all of that into the rules ahead of time.

So is there any way to exploit the potential communications space outside of frequencies, while at the same time protecting existing systems? I think there is. It requires us to borrow from a completely different domain of law: tort liability. Torts are injuries or accidents. If you punch me in the nose, I can sue you based on a tort theory. Tort doctrine balances your right to swing your arm against my right to avoid a bloody nose. No government agency decides ahead of time how either of us can act or grants exclusive rights to perform certain actions.

A tort-like system could be used for wireless communication. Every owner of a wireless device would have a baseline fundamental right to communicate, in any way he or she desired. Others could sue the owner for damages if his or her transmissions caused harm to their own communications. The legal system would balance the interests involved to determine who was actually at fault. The parties would have incentives to take appropriate care and factor the risks before engaging in any form of transmission. Technical standards or other baselines could be set as "safe harbors" to protect against liability if followed. This system could be implemented alongside the existing frequency-based allocations, creating new communications capacity without undermining legal obligations and business expectations.

With advances in wireless technology, any set of rules that declares some forms of transmission acceptable and others unacceptable will be inefficient. Something that interferes today may not interfere tomorrow, as devices get smarter and more adaptive. So let’s eliminate the rules, allow everyone to transmit however they wish, and use a liability system to deal with conflicts.

By killing off the illusion of a concrete spectrum resource, we would open up massive new opportunities for wireless communication.