Powering the Mobile Revolution: Principles of Spectrum Allocation

A mobile revolution is profoundly changing information technology. A dramatic shift from old technology norms built around fixed-location systems to new mobile devices, wireless networks, and location-centric applications has the potential to reshape society and the economy. In 2011 more smartphones were sold than personal computers, and by 2015, tablets will outsell personal computers as well. The new mobile technology increases economic efficiency and enables innovation, creating jobs and increasing prosperity. By the FCC’s estimate, mobile technology is “creating 1.5 million new U.S. jobs and offering tremendous potential to improve education, health care and public safety.”

The continued progress of the mobile revolution depends on advances in semiconductor technology, software, and spectrum reallocation. The most important of these inputs from a policy perspective is spectrum; the United States faces a “looming spectrum crunch” unless regulators can find 500 megahertz (MHz) of new spectrum for mobile applications in the next ten years to complement smaller cells and continued advances in signal processing technologies. Progress toward the 500 MHz goal has so far been mixed, as regulators oscillate between forward steps and backward ones. In order to be effective, legislators and regulators require a comprehensive set of spectrum management principles to guide spectrum policy decisions.

The United States leads the world in the adoption of 4th Generation LTE technology, but it lags the rest of the developed world in repurposing spectrum from legacy systems to LTE. Breaking the spectrum logjam depends on more liberal spectrum policies, better informed principles of spectrum rights, a better understanding of spectrum technology, and a great deal of hard, detailed work on the management of specific spectrum bands.

Nations that lead in the race to modernize spectrum policy stand to win the economic race for the jobs and innovation that are directly created by the new technologies, obviously, but they also stand to improve their economic performance a second time by replacing wasteful and antiquated information systems with more robust, pervasive, and efficient ones. In other words, they gain once by doing things they’ve never done before and again by doing the old things better.

Historically, spectrum allocation has been an ad hoc, piecemeal system driven by the logic of the moment: A commercial enterprise or government agency with an idea requested a spectrum allocation from the relevant regulator (the FCC in the case of commercial use, and the NTIA for government agencies). If the regulator saw merit in the idea, the regulator looked into its inventory of unassigned radio frequencies and allocated the best available fit. In some cases, the process of spectrum assignment has been initiated by the regulator itself; sometimes to good effect (Wi-Fi™) and sometimes not (satellite phones).

It’s now clear that spectrum allocation and management is an ongoing process that will benefit from guidance by a set of fundamental principles. Rather than simply re-assigning spectrum from legacy systems to mobile networks, policymakers need to reform the system that has created a critical shortage of spectrum in the most dynamic sector of the economy while over-allocating spectrum to wasteful and obsolete systems. The spectrum crisis is an opportunity for fundamental reform in the logic of spectrum assignment.

A more rational system of spectrum assignment would respect the principles that are evident in the operation of actual high-demand, high-performance, and high-efficiency wireless networks. In brief, these principles are:

1. Sharing: Prefer assignments that serve multiple users, as commercial networks do, over those for single uses.

2. Application Flexibility: Prefer assignments that support a variety of applications over those that support a single application.

3. Dynamic Capacity Assignment: Prefer networks that allow capacity to be adjusted on demand to those that allocate capacity statically.

4. Technology Upgrade Flexibility: Permit technology upgrade without permission.

5. Aggregation Efficiency: Prefer large allocations over small ones to minimize guardband losses.

6. Appropriate Facilities-Based Competition: Seek an ideal number of networks, a number that is likely to be larger than two and smaller than six in most instances.

7. High-Performance Receivers: Favor systems of high-performance receivers over those that can’t tolerate common sources of RF noise.

8. All Relevant Dimensions: Allocate “patches” of spectrum by frequency, power level, place, transmission direction, beam spread, modulation, coding, and time.

9. Promotion of New Technologies: Use rules modification rather than exclusive allocation as a means of enabling the next generation of spectrum technologies.

10. Maximize Redeployment Opportunities: When upgrades to existing systems free up spectrum for new ones, as was the case in the DTV transition, require the upgrade.

These allocation principles flow from empirical knowledge of the nature of spectrum, the current state of the art in radio engineering, and the likely timeline of new developments in radio engineering. They are explained in more detail in the main text.

Application of these principles to spectrum allocation disputes will help resolve case-by-case disputes in an optimal manner. Ideally, we should be able to score each spectrum dispute according to the number of principles it follows. This method enables us to determine the extent to which regulators are moving spectrum policy forward or backward. The examination of selected current controversies illustrates this method of analysis at work.

The demand for spectrum is largely created by new wireless applications. The most important example of the demand is the vast pool of applications that have been created for smartphones and intelligent infrastructure such as the “smart grid.” Demand for wireless data capacity—bandwidth—roughly doubles year after year.

Bandwidth is often compared to highway capacity, but a better analogy is food production: We can always build more roads, but we can’t increase the supply of arable land or of spectrum. We increase the food supply by bringing more acreage into agricultural use, by improving agricultural technologies such as genetically engineered seed, chemical fertilizers, herbicides and pesticides, and by employing sound management practices. Similarly, wireless bandwidth is increased by putting more patches of spectrum to use, developing technologies that increase bits/hertz usage efficiency, and managing network traffic responsibly. Each of these three practices is necessary, and each produces widespread societal benefits.

Spectrum research and development is extremely important, but in the short to medium term technology is not going to resolve the spectrum crunch on its own. Research is advancing along two principal lines:

1. Researchers on the Software-Defined Radio/Cognitive Radio (SDR/CR) field are developing techniques that allow easier access to unused or lightly-used patches of spectrum. These techniques are an alternative or a supplement to traditional regulator practices that assign spectrum to license holders who may not use their allocations fully at all times. In practice, SDR/CR needs to be connected to an authorization database such as the White Spaces Database that provides go/no go information to prospective network operators, and the decisions that his database implements flow from a spectrum allocation policy.

This branch of research is frequently touted as increasing spectrum efficiency, but this description needs clarification. SDR/CR actually aims to improve allocation efficiency by enabling a larger pool of potential users to contend for access to the spectrum. While this can be beneficial, it does not improve usage efficiency, the amount of information per unit of spectrum (bits/hertz) that can be transmitted and received over a given patch of spectrum.

2. Research on spectrum efficiency develops techniques that allow for greater bits/hertz usage efficiency. This line of research concentrates on techniques that govern the ways that bits are represented on wireless networks, the nature of antennas, and the coding systems and scheduling systems that enable multiple users to share a given patch of spectrum in an orderly manner.

Most of the practical advances in the use of RF spectrum by commercial and other public systems are the result of research on usage efficiency: Packet radio, modulation systems such as Orthogonal Frequency Division Multiplexing (OFDM) and Quadrature Amplitude Modulation (QAM), Multiple-Input Multiple-Output (MIMO) antenna systems, and scheduling/coding systems such Code Division Multiple Access (CDMA).

However, spectrum research doesn’t absolve policymakers from identifying more spectrum for wireless data systems. To the contrary, SDR/CR technology depends on spectrum allocation policy as it is primarily a means of implementing complex, multi-level allocation policies. Advanced research on usage efficiency is not currently mature enough to make allocation decisions unnecessary, even if it may be someday. If and when that were to occur, policymakers would still be required to address the vexing problem of the billions of less advanced systems that remain on the air with a transition plan to better technology.

Consequently, lawmakers and regulators concerned with spectrum allocation have no choice but to meet the current spectrum crisis by making better use of current technology. This requires repurposing and reallocating the pool of spectrum best able to meet the needs of the mobile revolution, a job that is best undertaken by adopting principles that reflect the best understanding of spectrum usage technology as it is today and as it will be in the next five to ten years.

Recent spectrum controversies don’t always reflect clear and consistent decision making. Grading the decisions that have been made or will be made soon in this area yields the following results with a scoring system that ranks each decision on a scale ranging from +10 to -10, where +10 is most desirable:

• Leave DTV Channel 51 live while imposing interoperability on 700 MHz B and C block license holders: -8
• Take DTV Channel 51 off the air: +8
• Remove 800 MHz internal guard bands: +10
• Convert government fixed point microwave to fiber backhaul: +10
• Replace Military Tactical Radio Relay with fiber: +10
• Move government video surveillance to a commercial carrier: +6
• Adopt PCAST spectrum sharing recommendation: +5
• Reassign LightSquared spectrum to wideband GPS: -1
• Adopt FCC Medical Body Area Networks plan: 0
• Allow Verizon to purchase SpectrumCo licenses: +7

Policymakers should generally strive for decisions that earn six points or more, and avoid decisions that earn less than three points in the absence of extenuating circumstances not captured by our grading system.

The spectrum agenda needs to proceed along two parallel time lines: In the short term, policymakers need to make more spectrum available for use by high-demand applications such as mobile broadband, and for the long term they need to support basic and applied research on spectrum to relieve capacity constraints. There is no downside in assuming that the spectrum crunch is real and that the long-awaited technical advances that promise to resolve it will not arrive for a very long time. There is an enormous potential downside in assuming that a technology solution that does not require re-allocating spectrum is just around the corner, however. The prudent course is to deal with today’s problem today while actively supporting the technology that we will use tomorrow.