Brought to you as a public service of the Open Spectrum Foundation (Stichting Open Spectrum), Amsterdam - Prague
This text is also available in Chinese: 什么是开放式频谱？
What is Open Spectrum?
"Open Spectrum" is based on the realization that technology can reduce or even eliminate the need for governments to micro-manage wireless communication. In different contexts it can be viewed as
For almost a century, governments have imposed detailed limits on the use of radio - who can transmit or receive what frequencies and waveforms, at what power levels, in which locations, for what purposes. Licenses summarize these controls for specific users or "stations." State control of radio usage goes far beyond what is accepted for other media, (speech, publishing, photography, Internet, etc.). Yet most people accept strict rules for radio in the belief that they are necessary to prevent chaos and interference.
However, during the past 20 years, smarter radios have been developed that go a long way toward solving problems which once seemed to require government intervention. Cordless phones can automatically scan a band to select an unoccupied channel. Cellular GSM phone networks dynamically assign frequencies when handsets are activated, and set signal levels to the minimum needed for an adequate link. Smart receivers can separate signals that are coded differently even when they occupy the same channel. Smarter radios tend to combine ease of use with better link quality and support for novel applications. The combination of these attributes has fueled explosive growth in public demand for wireless devices. And the spread of these devices dramatically improves economic efficiency, productivity, personal safety, convenience and social cohesion.
But the wireless boom also drew attention to the fact that regulations designed to protect "dumb" radio equipment from interference create artificial shortages of frequencies. Recent surveys have shown that static frequency assignments can result in band utilisation rates as low as 5-10 percent. A few radio experts began making this point in the mid-1990s, laying the groundwork for Open Spectrum to emerge as an alternative model in spectrum management. Eli Noam, Yochai Benkler, Dave Hughes and Kevin Werbach were among the first to argue for Open Spectrum as a serious policy option. But it was the US Federal Communication Commission's 1985 decision to allow new communication technologies in the bands for unlicensed "Industrial, Scientific and Medical" (ISM) devices that jump-started this evolution.
Communication in the ISM bands must tolerate interference. This is in contrast to traditional spectrum management, where the aim is to prevent interference. Protection against interference is normally achieved by not letting other transmitters use a licensed channel within a geographic "protection zone." But Wi-fi - a technology that developed in the ISM bands - showed that large numbers of people can share a band, without specifically assigned channels, if everyone uses low power and waveforms designed to soften the effects of interference. With no protection zone, there is no technical justification for licensing Wi-Fi. And indeed, most countries now exempt Wi-Fi from licensing, as shown by our global survey.
Wi-fi is often cited as Open Spectrum's "proof of concept," validating "unlicensed commons" as a practical paradigm in frequency management. That is why we focus on the de-licensing of Wi-fi as a confidence-building measure for regulators just starting on the path toward freer public access to radio. However, it is also important to note that Open Spectrum is a much broader concept than Wi-fi. At the same time, Wi-fi works as well as it does because of widespread voluntary acceptance of the IEEE 802.11b standard, and because of mandatory processes of "type approval" (in which equipment is approved by regulators for unrestricted sale if it conforms to certain parameters, particularly as to radiated power and frequency use). Thus, unlicensed is not the same as unregulated. Open Spectrum supporters seem split by this distinction, with some arguing for complete deregulation, and others (like ourselves) embracing type approval as preferable to licensing.
Another divisive issue is unlicensed access to licensed bands. Ultra-WideBand signals, for example, can be used at such low powers - below what is already allowed for unintended emissions, in fact - that many people think UWB interference to licensed services is unlikely. On the other hand, some licensees, who paid millions of dollars for their frequencies, do not want unlicensed UWB to get access to the same frequencies for free. A different approach is called "cognitive radio." Cognitive devices recognize the presence of other signals in their environment, quickly re-tuning to unoccupied channels and vacating each channel as soon as the primary licensee starts using it. Dynamic, opportunistic access like this could push band utilisation to very high levels indeed. But would the risk of interference from cognitive radios really be minimal? This is a question that needs more testing under a wide range of conditions. Nevertheless, opening a licensed band to shared use by unlicensed devices is easier than completely clearing a band of licensed users. So if you think of Open Spectrum as something that can be introduced gradually, this is an important option.
"Passive" radio services - like radioastronomy and satellite downlinks - pose a special problem. Their earth stations do not emit signals that could be detected by cognitive radios, and they must be able to receive extremely faint signals without interference. It may be that the bands for passive services need to be protected like wilderness areas and excluded from unlicensed use. We have no problem with that - but other advocates of Open Spectrum think all radio bands should be unlicensed commons regulated only by "the minimum set of rules required to enable the success of the wireless commons."
Regardless of the problems and risks, some people think radio technology is evolving inevitably toward a future where traditional forms of regulation will be impossible. Billions of Radio Frequency IDentification (RFID) tags are likely to spread around the world in the coming decade; they will be as hard to control as an epidemic. "Software-defined radio" is another challenge. More and more radio functions that had been performed by hardware are likely to be implemented in software in the future. If this software is open source, or can be modified or replaced after purchase, "type approval" processes are undermined.
What then? Optimists envision a post-regulatory future where Darwinian competition in the marketplace will yield the "fittest" devices - equipment immune to interference and capable of automatically finding and exploiting any morsel of under-used spectrum when needed. Just as the largest animals tend to be placid vegetarians, the users of powerful radio equipment may choose not to cause interference, for the benefit of all. Widely supported standards and protocols for automating co-operation and "politeness" seem essential for the success of this scenario. If some spontaneously self-organizing system can be devised for quickly imposing penalties on interferers, everyone could feel more confident in such a future.
Our own goals are more modest. We would like to eliminate governments' role in granting permission for individuals and organizations to use radios harmlessly, particularly in developing countries. To state that positively, we want to unleash the benefits of wireless communication for economic and social development in the places that have the most to gain. As the need diminishes for treating radio with special strictness, to counteract the technology's shortcomings, radio regulation should converge toward the rules that apply to the most common medium - ordinary human speech. It may take a while to get there, but progress along the way will be marked by economic growth and improvements in the quality of life.
---Robert Horvitz, 29 June 2005