The End of Spectrum Scarcity

Radio spectrum may be one of the most tightly regulated resources of all time. From cellphones to police scanners, from TV sets to garage-door openers, virtually every wireless device depends on access to the radio frequency wireless spectrum. But access to spectrum has been chronically limited ever since RF transmissions were first regulated in the early 20th century. Now that's all about to change. New technologies that use spectrum more efficiently and more cooperatively, unleashed by regulatory reforms, may soon overcome the spectrum shortage.

Since the 1920s, regulators have assumed that new transmitters will interfere with other uses of the radio spectrum, leading to the "doctrine of spectrum scarcity." As a result, every wireless system has required an exclusive license from the government. With virtually all usable radio frequencies already licensed to commercial operators and government entities, the upshot has been, in the words of former U.S. Federal Communications Commission (FCC) chair William Kennard, a "spectrum drought." We've become accustomed to seeing every new commercial service, from satellite broadcasting to wireless local-area networks, compete for licenses with numerous existing users, including the government--all of which guard their spectrum jealously. Cellular phone service, for example, was demonstrated in the lab in 1949 but not deployed until the 1980s, largely because of licensing delays.

That world is coming to an end. At least in the United States, new technologies and regulatory reforms may soon free up enough RF capacity to transform wireless-industry economics, especially for popular mobile telephony and wireless Internet services. In fact, there's every reason to think we're on the cusp of a spectrum explosion--one that will trigger major shifts in investment, business models, and services.

In the spectrum-rich future, wireless connections for new voice, music, and video services should abound, benefiting consumers and businesses alike. In our homes, devices such as TVs, stereos, DVD players, and PCs will come with built-in high-capacity wireless links to swap information. Outside, new networks will let movies and other huge multimedia files zip across town or across the country. Billions, or perhaps trillions, of wireless sensors will be embedded virtually everywhere. Wireless data, voice, and video connections will be increasingly available when we are on the move, in cars, trains, and perhaps planes, too. New services--everything from personal music channels to video-on-demand to mobile computing utilities and, yes, to the latest in ever-profitable adult entertainment--will flourish. So, too, will the markets for the hand-held devices needed to deliver these services.

These scenarios do not require infinite bandwidth. Relatively modest capacity increases--from either new spectrum allocations or new technologies--can have dramatic consequences. Today, satellite radio is delivering scores of new music choices to millions of listeners nationwide using just 25 megahertz of spectrum, about the same bandwidth as four analog television channels. Personal communications services have sparked a sea change in data services delivered to cellphones, using about 90 MHz. The Wi-Fi (the popular name of the IEEE 802.11 standard) revolution in wireless local-area networking was started with only 84 MHz. Now imagine more new spectrum made available simultaneously in the next few years than is now used by the satellite TV, PCS, and Wi-Fi industries combined [see table, " "].

The era of future abundance will be as foreign to us as our world today would have been to Marconi and Tesla, whose early spark-gap radios occupied the entire usable spectrum for each individual Morse code message. The U.S. government's first tables of spectrum allocation, in the 1920s, extended only to 60 MHz, with frequencies above 23 MHz labeled "experimental." The bands of spectrum covered by international treaties were similarly limited. In contrast, our current allocation tables regulate spectrum up to 300 000 MHz (300 gigahertz), with the vast majority of services operating above the 60 MHz that was once the top of the chart, beginning with the FM radio band (88­108 MHz).

Before we look further at what it means to live in an age of spectrum abundance, let's look more closely at the two main reasons for the past era of scarcity: the state of available radio technologies and government policies. What's extraordinary about the present period is that both these historical constraints are simultaneously going through radical change. Let's start with technology.

To understand the impact new radio technologies are having on spectrum availability, it is helpful first to address a common misconception: that spectrum is a concrete and finite resource. Not so. Radio waves do not pass through some ethereal medium called "spectrum"; they are the medium. What's licensed by governments is not a piece of a finite pie but simply the right to deploy transmitters and receivers that operate in particular ways.

Moreover, interference is not some inherent property of spectrum. It's a property of devices. A better receiver will pick up a transmission where an earlier one heard only static. Whether a new radio system "interferes" with existing ones is entirely dependent on the equipment involved. Consequently, the extent to which there appears to be a spectrum shortage largely depends not on how many frequencies are available but on the technologies that can be deployed. Many regulations intended to promote harmony of the airwaves have instead, by putting artificial limits on technology, created massive inefficiency in spectrum utilization.

Last year, a Spectrum Policy Task Force, organized by the FCC, recognized that much of the spectrum already licensed is not really in short supply. If you scan portions of the radio spectrum, even premium frequencies below 3 GHz in dense, revenue-rich urban areas, you will find that most bands are quiet most of the time. One study found that only four of 18 UHF television channels were used in Washington, D.C. Sometimes that's by design, as with "guard bands"--spectral equivalents of highway shoulders, in which no radio signals are permitted. Fifty years ago, when TV sets still used vacuum tubes, guard bands were the only way those sets could distinguish signals on adjacent channels. In some other cases, an apparent lack of spectrum use reflects system design, as with cellular-phone towers, which transmit actively only when communicating with a nearby handset.

That's why what's happening now is so exciting. New radio transmission and networking technologies can squeeze more and more capacity out of the same spectrum. Some of the improvement comes from the shift from analog to digital transmission. For example, at least five digital TV shows can be broadcast on the same frequencies that a single analog channel now occupies. Similarly, digital cellular systems now carry three times as many phone calls as their analog predecessors.

Even greater improvements in spectrum usage will come from a family of technologies that use the computational intelligence of today's wireless devices to allow multiple systems to "share" the same spectrum. The first of these, spread spectrum, replaces ancient high-power, undifferentiated narrowband transmissions with modern low-power, coded wideband signals [see figure, " What's the Frequency?"]. First described during World War II, spread-spectrum technology is already used in many cellular phone networks and in Wi-Fi, but newer systems promise even greater capacity improvements.

A newly permitted method of using spectrum, ultrawideband, takes spread spectrum to its logical conclusion, operating at such low power that, subject to appropriate safeguards, it can underlie existing licensed services. That is, preexisting users of the same spectrum bands won't even know the ultrawideband transmissions are there. It will be as if we figured out a way for freight trains to travel on highways, with cars being none the wiser. Standards work is already under way to make ultrawideband the core technology for home entertainment networks, transferring video, audio, and photos among home PCs, stereos, high-definition televisions, and DVD players.

And this is only the beginning. Another recent innovation, smart antennas, can focus adaptively to "lock into" a directional signal. Instead of radiating a signal in all directions equally, they figure out where a user is located and direct the radiation accordingly, reducing effective interference with other transmitters. Now, too, novel coding algorithms can take factors that traditionally hampered transmission, such as physical obstacles and motion, and use them to generate information that increases capacity.

Perhaps the greatest technological gain in wireless capacity, however, will come from systems that work cooperatively. In a network architecture called a mesh , each RF receiver also acts as a transponder, retransmitting data sent by other devices in the network. In other words, every new device uses some of the network's capacity but also adds capacity back. Because a device in a mesh no longer needs to send information all the way to its ultimate destination (such as a cell tower), it can use less power. That allows the network to add more devices without any noticeable increase in interference. The approach resembles the distributed architecture of the Internet, in which every router can move traffic along an efficient path.

Software radios are a key enabler for all these advances. A software radio can receive and transmit across a broad range of frequencies; because it processes signals in software, it is far more adaptable than a traditional radio. In principle, a software radio originally used for cellular telephony could, for example, download new software and begin to receive broadcast television signals, or, more likely, access a network that uses a new cellular transmission protocol. Even more sophisticated "cognitive radios" would work cooperatively, analyzing other nearby radios and adapting on the fly to avoid other transmissions.

The spectrum "dividends" possible from these new technologies have not been lost on regulators. Traditional spectrum licenses were technology- and service-specific, precluding most of the capacity-enhancing mechanisms described above. Led by FCC chair Michael Powell, the U.S. government has embarked on a historic effort to update the way spectrum is managed. It has three main strands:

Spectrum reallocation: the reallocation of bandwidth from government and other long-standing users to new services, such as mobile communications, broadband Internet access, and video distribution.

Spectrum leases: the relaxation of the technical and commercial limitations on existing spectrum licenses by, for example, permitting existing licensees to use their spectrum for new or hybrid (for example, satellite and terrestrial) services and granting most mobile radio licensees the right to lease their spectrum to third parties.

Spectrum sharing: the allocation of an unprecedented amount of spectrum that could be used for unlicensed or shared services.

Spectrum reallocation The FCC's reallocation of 120 MHz of spectrum for third-generation (3-G) mobile services has probably received the greatest media attention. It stemmed, in part, from a landmark 2002 agreement with the U.S. military to free at least 45 MHz of government spectrum. What's more, as part of the digital television transition, the FCC is reclaiming and auctioning approximately 85 MHz of UHF broadcast spectrum, which might be used for mobile communications services in the future.

Once TV stations commence all-digital broadcasting on their newly assigned channels--perhaps as early as 2007--the FCC will also reclaim a huge cache (approximately 290 MHz) of "beachfront" spectrum, assigned for analog broadcasting since the 1940s. Germany, which switched off analog broadcasting in Berlin this year and is now enjoying smooth-running all-digital broadcasting, shows just how fast this transition can occur.

Since 1999, the FCC has also authorized more than a dozen new high-powered satellite systems to use at least 500­1000 MHz of spectrum for broadband video and data services in the 26.5­40-GHz band (known as the Ka band). And in 2000, the commission authorized the reuse of 500 MHz of spectrum for a novel tower-based video and data service that shares spectrum with existing satellite television.

The spectrum portfolios of large cellular phone companies will certainly be devalued

This new scheme, first envisioned by Northpoint Technology Ltd., of Portsmouth, N.H., and known as the multichannel video distribution and data service (MVDDS), takes advantage of the fact that in the northern hemisphere, satellite TV dishes all face south--the satellites themselves orbit over the equator. So the service would send customers about 100 channels of digital television programming--and throw in high-speed Internet access as well--by transmitting southward to north-facing dishes from lower, tower-based radios.

Taken together, these and other FCC actions may increase the available bandwidth for video services by over 1500 MHz and the bandwidth for mobile wireless services by at least 300 MHz. By comparison, from 1985 to 2000 the total allocation for analog and digital cellular telephony, plus specialized mobile radio services, was just 195 MHz.

Spectrum leases These spectrum reallocation decisions are complemented by a May 2003 FCC decision to relax the conditions under which spectrum can be leased to another party. Before then, licensees could neither sell nor "sublet" their spectrum to another party.

Economists have long urged the FCC to make spectrum licenses more flexible and to authorize secondary markets through which licensees could lease some or all of their rights to others who could use the spectrum more efficiently. Such changes also would get the FCC out of the impossible task of deciding ahead of time what spectrum is worth and how it should be used.

The agency's May 2003 order endorses this idea, but only in part. It permits most two-way wireless communications services, such as cellular voice and data licensees, to lease some or all of their spectrum. But UHF operators and other broadcasters are not covered by the new rules, even though most licensed UHF television channels are simply not used. In addition, the wireless licensees that are covered can lease their spectrum only for services of the same type as their own. A group of paging operators, for example, probably could not pool their spectrum and lease it for two-way mobile telephony.

Spectrum sharing The dramatic growth of Wi-Fi in the unlicensed or shared 2.4-GHz band has been one of the few unequivocal telecommunications success stories in recent years. From fledgling sales in 1999, the market for Wi-Fi equipment grew to US $2 billion in 2002 and will exceed $4 billion in 2005, according to a projection by In-Stat/MDR, in Scottsdale, Ariz.

The unlicensed bands are not unregulated. The FCC still sets aside specific frequencies for unlicensed use, imposes power limits on transmission equipment, and provides some basic technical rules for receivers. However, once these requirements are set, any company can build and deploy compliant radio equipment, for any use, without needing further permission from the FCC or a spectrum licensee. A combination of sophisticated coding techniques, powerful signal processing, and widely adopted technical standards allows a cacophony of unlicensed devices, from Wi-Fi to garage-door openers, to coexist without any operator or government deciding what can be deployed.

Wi-Fi's success helped convince the United States to support the worldwide addition of 255 MHz of unlicensed spectrum in the 5-GHz range and to persuade other governments at the 2003 World Radiocommunication Conference in Geneva, Switzerland, last summer to make this new allocation a global reality. As a result, in the years ahead, the 5-GHz band could well provide a home for more high-speed Wi-Fi devices (which use IEEE Standard 802.11a, as opposed to 11b or 11g). That same band could be used by new metropolitan-area networks, which offer Internet access from cellular-like towers [see "The Wireless Last Mile," IEEE Spectrum, September 2003].

In the future, the type of spectrum sharing that the FCC permits in unlicensed bands may also be possible in licensed bands through the use of spectrum underlays and cognitive radios, described above.

The technical and regulatory developments surveyed here suggest that we are beginning an entirely new era--an era of spectrum abundance. Because these changes have largely occurred piecemeal, few participants seem to have recognized their collective impact. Every segment of the communications industry cares intensely about its own spectrum allocations but pays less attention to policy developments affecting other frequency bands and services.

To be sure, not all spectrum is created equal. The commercial value of any frequency band depends on its propagation characteristics and the current deployment environment. But, in today's world of digital convergence where almost every electronic service can be linked to via the Internet, most customers will be unconcerned which band of RF spectrum is used to deliver the content they want or to provide communications. Bits are bits, and spectrum is spectrum, as long as it connects.

Clearly, in the decade ahead, spectrum access will become more like a commodity--widely affordable for almost any purpose the user desires. While it may be too early to identify specific winners and losers, we can venture some general observations about the impact of a post-scarcity regime.

When spectrum seems scarce, service providers must focus all their energy on acquiring and husbanding spectrum access rights. But when spectrum seems more like a commodity, investors will shift their funds from those that are best at protecting spectrum to those best at using it: content and service providers. Likewise, economic power will shift from government spectrum gatekeepers to consumers, and from companies able to obtain licenses to those that offer useful devices, content, and service packages.

Incumbent mobile operators and broadcasters will almost certainly face greater competitive pressures from both licensed and unlicensed alternatives. The spectrum portfolios of incumbent operators, especially the large cellular phone companies, may be the first to be devalued. Manufacturers, on the other hand, may see an enormous stimulus from the new spectrum environment. If nothing else, lower entry barriers mean that more service providers will want their equipment. Greater demand, in turn, may stimulate price reductions for devices and other equipment.

As spectrum becomes more abundant, content providers will benefit, too. Indeed, consumers seem to have always-open wallets when it comes to new mobile applications. Cellphones are the device of choice for almost everything now--music, AM/FM radio, and mobile games, especially interactive ones, and even Web-based news, sports, and weather.

We really don't know yet how the coming age of spectrum abundance will reshape the market for communications services. But it's tempting to speculate that as spectrum becomes ever cheaper, the wireless data and mobile phone markets will begin to look more like the computer industry. Both, after all, benefit from Moore's Law.

In this new world, services may be driven more by the cycle of new receiver technologies than by access to spectrum licenses, more by engineers and entrepreneurs than by lawyers. The broadcasting business might change in a similar fashion, with content becoming more like network applications, available for a price anytime and anywhere, rather than at a fixed time on one channel. And in a world of spectrum abundance, regulators will be much less important. Spectrum users themselves will play a bigger role in mitigating potential conflicts and in putting spectrum to its highest and best use. At long last, the epoch of spectrum scarcity will be over.