Unlicensed Spectrum: The Challenge and the Opportunity

This is a story about a uniquely successful government initiative. Over the last three decades, the Federal Communications Commission has created and expanded bands of spectrum which are made available for unlicensed use. Unlicensed use, particularly for Wi-Fi, has enabled a proliferation of consumer devices such as garage door openers, remote controls, baby monitors and wireless speakers. There are also innumerable vital industrial and scientific applications using unlicensed spectrum. A new analysis from the Consumer Electronics Association estimates that unlicensed spectrum uses already generate $62 billion annually for the U.S. economy. Even though this may be somewhat overstated, there is no doubt that freeing up spectrum for unlicensed use has been a roaring success.

While many in the tech community, including public interest groups, want even more spectrum to be set aside for unlicensed use, incumbent industries, such as wireless phone companies, prefer the widest possible use of the current system of exclusive licenses.

A warning to techies and communications lawyers: to focus on the policy issues surrounding unlicensed spectrum, it is necessary to simplify, and sometimes, oversimplify, the technical and legal framework for spectrum allocation. Purists need not read further.

First, some background. The basic scheme of the Radio Act of 1927, and the Communications Act of 1934 which followed it, is that swathes of spectrum were allocated to particular uses (e.g., television, radio, radar, public safety, etc.), and portions of those bands were then licensed on an exclusive basis to a particular user, who would serve as a licensee for a fixed length of time. Licenses were free of charge. Then, as now, licenses are awarded and renewed under the FCC’s “public interest standard,” which is a broad mandate to require that our communications networks serve the “public interest, convenience and necessity.” Although, in theory, these licenses are subject to non-renewal, and may be revoked in extreme cases, in practice, licenses are invariably renewed. Licensees are protected by the government; it is a crime to interfere willfully with licensed transmissions.

Licenses can be sold and often are for large amounts of money. (Supposedly, what is sold are the assets of the licensee, not the licenses themselves, but purists are reminded of the warning above.) Thus, although licenses do not confer ownership of the spectrum, those who hold them have what they generally consider to be tantamount to a property right in the spectrum.

The original system worked quite well for a very long time. However the selection process for exclusive licensing, which used “beauty contests” pursuant to the public interest standard, became increasingly slow, wasteful and anachronistic. (Many, including this author, still think that the old system should be retained for broadcast TV and radio.) For a brief period in the 1980s, the Commission resorted to selecting licensees by lottery, but this proved to be an even worse solution.

Meanwhile, in 1959, economist Ronald Coase proposed that spectrum should be auctioned. While this argument was pooh-poohed by most policymakers, it gradually gained traction and, by the 1980s, the FCC was asking for authority to conduct spectrum auctions. Eventually, in 1993, the FCC was first granted authority to conduct certain spectrum auctions; this power was broadened over the next several years. Today, with a few exceptions, auctions are the rule.

Although many libertarians advocate the outright sale of spectrum ownership rights, the FCC now auctions licenses for a fixed term (e.g., 15 years). This gives the Commission flexibility for the future, and greater control over interference and other technical considerations.

The central tenet of spectrum licensing, reinforced by the fact that they are allocated by auction, is its exclusivity. While this system continues to work well in many instances, and is, indeed, necessary under current conditions, technology has made it possible to use spectrum much more efficiently, through various non-exclusive sharing mechanisms. In some cases, licenses can be modified so two or more users can use the same spectrum. Down the road, it will ultimately become possible to create intelligent “cognitive radio” devices that will allow far more efficient use of spectrum (by orders of magnitude), but even now, the greatest potential for efficiency lies in unlicensed uses. (There is another category, not relevant here, called “licensing by rule.” Don’t ask.)

In the eyes of many people in the tech community, the quasi-property based exclusive licensing system is too restrictive. Under an unlicensed scheme, bands of spectrum may be used by anyone, for any purpose, as long as they adhere to a few rules. Most importantly, their devices must not use too much power, which means that they have very little range, but are not likely to interfere with other users.

Going back as far as 1938, the FCC had allowed very low power devices to operate without licenses, but manufacturers needed FCC permission for each category of use. In any event, given the limitations of technology, these devices were not very important in the grand scheme. In the 1980s, some far-sighted engineers at the FCC, including Michael Marcus (who is generally credited as the moving force), had a different idea. Under the new scheme, the FCC said that all low power uses in a number of bands of spectrum would be permitted. Manufacturers would need FCC certification to insure that devices would only operate in the permitted bands and at the permitted power, but users were then free to do anything they wanted with them. The flip side is that if someone else’s device caused interference, users would have to live with it.

Among the bands that the FCC opened to unlicensed use is the 2.4 Ghz (gigahertz) band, which was primarily used for industrial scientific purposes. Spectrum engineers derisively belittled the 2.4 Ghz band as “junk spectrum,” because they deemed it unlikely to have any meaningful applications. (Today, that band is used for Wi-Fi and all manner of other uses.)

But technology was catching up. Moore’s Law, which hypothesizes that computer processing power doubles approximately every 18 months or two years (depending on whose definition you use), fueled rapid advances in networking technology and dramatically reduced costs. Thus, as the Internet evolved, unlicensed uses, especially on the 2.4 GHz band, became more and more feasible and attractive. The FCC opened portions of another important band (above 5 Ghz) for unlicensed uses, and, by the end of the century, the development and sale of unlicensed devices began to soar.

The philosophical father of modern unlicensed use is Yochai Benkler of Harvard Law School. In an influential 1998 article, he proposed that wireless networks, not just individual devices, could operate without the need for exclusive licensing. This, he said, could confer the value of the spectrum upon individual users rather than licensees. Property law (which is, for immediate purposes, the driving force of exclusive licensing) relies on exploiting scarcity to create value. Benkler argued that greater value (and greater freedom of speech) can be derived through ubiquity, making spectrum available for all to share and use. He said that technology would soon transcend the need for exclusive licensing. While most libertarians still prefer a private property scheme, some have come to view unlicensed spectrum as a desirable alternative, since it minimizes government involvement.

As a matter of technology, we are a long way from the point when exclusive licensing will be obsolete. However, the FCC has opened the way for expanded use of unlicensed spectrum through use of advanced technologies that can detect authorized or priority users and power down or shut off as needed. It is also implementing a system in which authorized spectrum users register on a real-time database, which unlicensed devices can consult to determine if there is room for them to transmit. This has enabled devices to access spectrum bands with very desirable propagation qualities, allowing longer range and improved building penetration.

The real challenge to unlicensed uses is not technological; it is political. Wireless operators like AT&T and Verizon, and equipment manufacturers like Cisco have hundreds of billions of dollars invested in the status quo. Their anxiety has only been magnified by the cable industry’s increasing use of Wi-Fi to create a competitive alternative for Internet access for their customers. Incumbent licensees, and some members of Congress, liken unlicensed spectrum to a government “giveaway,” presumably because the same spectrum could be sold at auction. Proponents take umbrage at this concept, since the spectrum has always been owned by the public, and it generates so much economic value.

An example of the fights generated when spectrum is at stake arose recently, when the unlikely duo of Senators Marco Rubio (R-FL) and Cory Booker (D-NJ) introduced legislation to speed the conversion of part of the 5GHz band to unlicensed use. (Interestingly, support and opposition to unlicensed spectrum cuts across party lines.) This spectrum has long been set aside for future use in developing “intelligent highways.” Predictably, the automobile industry’s Intelligent Transportation Society of America quickly cried foul, warning that “Motor vehicle crashes are the leading cause of death for children and young adults in the United States, with an annual death toll of 33,000 and over 2.3 million people injured on our nation’s roadways,” and that the Rubio/Booker legislation could impair efforts to use technology to reduce highway deaths.

Expanding unlicensed uses of spectrum is a politically charged endeavor, but it seems as if the promise of unlicensed technology makes the effort worthwhile.


By Andrew Jay Schwartzman.