page image

CHAPTER II - FCC Regulation of Spectrum—A Brief History

The federal government imposed command-and-control regulation of spectrum 90 years ago in response to conditions of overcrowding and interference in the AM radio broadcasting bands. Established by the Radio Act of 1927, the basic administrative model—block allocation and licensing based on “public interest” criteria—carried over into the Communications Act of 1934, and it is still in use today. Under this approach, the FCC allocates an individual band to one or more narrowly defined uses (“services”) and permits little significant modification of the designated use. In addition, it specifies the power limits, build-out requirements and other rules to which the service(s) allocated to such a band must adhere, based on the technology and business models that existed at the time of the allocation. Finally, prior to the mid-1990s, the FCC assigned licenses for individual bands using a non-market mechanism—historically, it employed a slow and costly comparative hearing process.

Alternatives to Command and Control

Although most prime spectrum is still subject to command and control regulation, in recent decades the FCC has embraced two alternative models: a market approach that treats exclusively licensed spectrum like private property and a spectrum “commons” approach that eschews licensing altogether.i These alternative models are not mutually exclusive, and both have yielded dramatic gains for consumers. Nevertheless, proponents of the two models have engaged in a long-running intellectual battle over their relative benefits.ii Recently, this lively and important debate has expanded to include a third approach, dubbed “spectrum sharing.”

Exclusive (Flexible) Use. The licensed alternative to command-and-control regulation preserves the benefits of exclusivity while taking advantage of flexible, transferable spectrum use rights. Economists have long favored a market approach to the allocation of resources, generally, and spectrum, in particular.iii As early as 1959, Nobel Prize-winning economist Ronald Coase wrote that spectrum was a fixed factor of production, like land or labor, and should be treated in the same way, with its use determined by the forces of the market rather than the decisions of government. In an analysis of FCC regulation that led directly to the seminal essay he published the following year, Coase concluded that the assignment of well-defined property rights in spectrum use that such an allocation would entail, would be sufficient to prevent inefficient broadcast interference.iv

Economists argue that a market-based approach to spectrum regulation has two advantages over command-and-control regulation (and over unlicensed allocation, discussed below). The first is efficiency in use. Economists believe that the profit motive will deliver spectrum, like any other valuable resource, to those who can put it to the uses most desired by the public. Over time, the inexorable pressure to make efficient use of a scarce resource such as spectrum leads to increased investment and innovation, which produces dynamic efficiency.v

Although the FCC traditionally allowed some role for the market in spectrum management (e.g., radio licenses have long been bought and sold), that role expanded markedly in 1993, when Congress authorized the use of auctions to award spectrum licenses for non-broadcast services. The FCC has auctioned 385 megahertz of “new” spectrum for mobile voice and data services (known collectively as Commercial Mobile Radio Services, or CMRS) since then, and it has taken additional steps to promote a secondary market in spectrum rights. Even more important, the FCC has granted increasingly flexible spectrum usage rights to CMRS operators, allowing them to choose which services to offer and the technology with which to deploy them.

This model of exhaustive assignment of exclusive, flexible rights has worked extremely well for CMRS carriers, who want predictably high quality-of-service and 24/7 availability over large geographic (including national) coverage areas. In 2013 alone, U.S. wireless carriers spent $33 billion to build out and upgrade their networks, a pattern of capital investment that has allowed them to seamlessly deploy successive generations of wireless technology.vi The fourth generation (4G) technology, LTE (for Long Term Evolution), is designed to transfer large amounts of data (e.g., video streaming) at high speeds and to operate in high-interference environments. These investments have generated enormous benefits for consumers: As one indication, in 2013, U.S. wireless carriers generated $189 billion in revenue.vii The steep prices paid in the FCC’s recent auction are another indication of the value of exclusive, flexible-use rights to spectrum.

Unlicensed Use. At the same time that market-based reforms of spectrum regulation were gaining wider acceptance, some legal scholars and a group of technology firms were urging the FCC to eschew licensing altogether and treat the spectrum, or large blocks of it, as a common resource. As early as 1938, the Commission adopted rules (Part 15) allowing for the operation of non-licensed, low-power devices that did not cause harmful interference to licensed services. For many years, most Part 15 devices were designed to operate below 30 MHz, but over time, the FCC amended and expanded the rules to permit such devices to operate at higher power in certain higher frequencies. In the 1980s, the FCC permitted the use of spread spectrum radio systems on an unlicensed basis at a significantly higher power level in three bands (902–928, 2400–2483.5 and 5725–5850 MHz).viii Over time, the rules governing unlicensed bands have been relaxed to permit the use of any digital modulation, not just spread spectrum, and in response to a growing chorus of supporters, the FCC has set aside significant additional spectrum for unlicensed use, including 425 megahertz in the 5 GHz band.ix

Proponents argue that the unlicensed model—by eliminating the cost of spectrum acquisition and infrastructure investment—enables device manufacturers and service providers to develop markets in sophisticated equipment and network services built on them to deliver reliable connectivity.x In response to challenges that the absence of licensing will result in a “tragedy of the commons,” proponents maintain that the imposition of exclusive-use rights to prevent interference is neither necessary nor desirable. Exclusivity is not necessary, they argue, because the amount of spectrum that many new uses require is minimal, and newer receivers can adapt to the presence of interference at levels that would have caused traditional technologies to fail. Moreover, it is in the interests of equipment manufacturers to invest in designs that perform robustly in the presence of interference as a way to expand their volume of business. Nor is exclusivity desirable, according to this view, because the transaction costs of arranging to get the necessary spectrum rights for this newer technology would exceed the benefits.

The unlicensed model has enabled a game-changing set of spectrum uses that is both different from and complementary to CMRS. The “killer app” is Wi-Fi hot spots—small, isolated base stations that allow wireless devices in a home or office to connect to the Internet through a wireless or wireline connection. Wi-Fi’s ability to operate independent of a larger wireless infrastructure, together with the availability of significant amounts of unlicensed spectrum, make possible low-cost devices that operate at extremely high data rates.

In addition to enabling Internet access in homes and businesses, unlicensed spectrum is key to the emerging markets in machine-to-machine (M2M) connectivity (also called the “Internet of Things”), including radio frequency devices for inventory management, automated meter readers and other “smart grid” applications, and wireless health care. Significantly, mobile broadband itself has become reliant on unlicensed spectrum, as CMRS carriers use Wi-Fi to offload an increasing amount of Internet-related data traffic from their congested (licensed) networks.

The barrier separating unlicensed Wi-Fi and licensed CMRS will become even less distinct once cellular operators implement their plans to extend LTE technology to unlicensed spectrum. Although this technology, dubbed LTE-U, could in principle be deployed in any unlicensed band, the CMRS operators are focusing initially on the 5 GHz band, in part because of the large amount of unlicensed spectrum (500 megahertz) that is available there on a global basis. LTE-U, also known as Licensed Assisted Access (LAA), will allow a CMRS operator to integrate the licensed and unlicensed elements of its network into a unified LTE network. Among other benefits, that will reduce the complexity of the hand off between Wi-Fi and CMRS. In addition, because LTE networks are managed, they use spectrum more efficiently than Wi-Fi networks, especially when demand is high.xi Although efficiency is not the only relevant goal for such networks—competitive access and resilience are also important—LTE’s ability to carry more data traffic in a fixed amount of spectrum will complement the ubiquity and other strengths of Wi-Fi in the unlicensed bands.

Despite this potentially powerful complementarity, the implementation of LTE-U/LAA raises important implementation issues that the LTE and Wi-Fi standards organizations are working together to resolve. A key issue is the nature and transparency of the algorithm used to allocate spectrum on a dynamic basis between LTE devices and Wi-Fi devices. Although it will take time to resolve these issues, LTE-U/LAA is almost certain to be deployed in the next few years.

Shared Use. Recently, the debate over spectrum management has expanded to include a third alternative to command and control—referred to as spectrum sharing, shared use or protected shared access. Spectrum sharing is nothing new. In exclusively licensed spectrum, a CMRS operator manages the sharing on behalf of its customers (“cooperative” sharing), and in unlicensed bands, opportunistic users engage in unmanaged (“non-cooperative”) sharing among themselves and between themselves and primary users. By contrast, protected shared access involves sharing among operators of multiple radio networks (not just CMRS), each with licensed, interference-protected access rights. Some versions of the model include opportunistic sharing as well.

This emerging model is a response to the fact that much prime spectrum is encumbered but underutilized. In particular, many prime bands are assigned to government users or other incumbents that cannot be relocated in the short-term or that have a long-term need for the spectrum but use it on an intermittent or piecemeal basis, e.g., only during certain time periods and/or in specific geographic areas. Largely due to advances in technology, it may be possible for regulators to allow a limited number of licensees to operate, on an interference-protected basis, in a frequency band that is already assigned to one or more incumbent users. Secondary licensees could be required to pay for (shared) access, which would create an incentive for incumbents to free up or share underutilized spectrum.

Proponents of this model have focused initially on higher frequencies (the 2.3 GHz band in Europe and the 3.5 GHz band in the United States), which lend themselves to the deployment of small cells. Cellular operators use small cells in part to augment capacity in places where the demand for network connectivity is highly concentrated (e.g., sports stadiums and transit stations).xii Because of their limited geographic coverage and low-power transmissions, small cells are less likely to interfere with incumbent systems, thus reducing the size of the “exclusion zones” needed to protect such systems.

The protected shared access model is a global development. In 2012, the European Commission asked the relevant organizations in Europe to develop standards to enable the Licensed Shared Access (LSA) model, also known as Authorized Shared Access (ASA).xiii LSA/ASA is a two-tier system for sharing licensed spectrum between incumbents and secondary users in the 2.3 GHz band, which is widely used outside Europe for mobile broadband but is encumbered inside Europe. That same year, in the United States, the President’s Council of Advisors on Science and Technology (PCAST) proposed something similar to LSA/ASA but with a third tier for opportunistic access. The PCAST proposal focused on the 3.5 GHz band, which is allocated to high-power military radars and non-federal Fixed Satellite Service (FSS) earth stations. More broadly, PCAST called for the creation of 1,000-megahertz-wide “spectrum superhighways” in which dynamic spectrum sharing would replace the single-use allocation as the normal mode of operations.

PCAST also called for the regulation of receivers. Although interference is a reciprocal harm—it results from the performance of the receiver no less than that of the transmitter—regulators have traditionally viewed transmitters as the source of interference and receivers as innocent “victims.” In some cases, this limits the potential for adjacent bands to support valuable new services. PCAST argued that having clearly defined “receiver interference limits” would help achieve more efficient trade-offs between the rights of transmitters and receivers, with the goal of maximizing concurrent operations as opposed to minimizing harmful interference.

Shortly after the PCAST report was issued, the FCC proposed to create a Citizens Broadband Radio Service (CBRS) in the 3.5 GHz band along the lines of the PCAST proposal.xiv Following a two-year proceeding, the FCC recently adopted rules that allow for commercial operation in 150 megahertz at 3550–3700 MHz using a three-tier sharing system.xv Two CBRS tiers of users—Priority Access (PA) and General Authorized Access (GAA)—will share the band with a third tier of protected incumbents (military radar and FSS users). The Report and Order sets aside 100 megahertz of the band (3550–3650 MHz) for PA use. Would-be PA users will be able to bid via auction for short-term, geographically targeted licenses that afford interference protection from GAA users. (The rules facilitate PA deployment of small-cell technologies, for which the 3.5 GHz band is well-suited.) An 80-megahertz block of channels will be reserved for opportunistic use by any FCC-certified GAA device. GAA users will also be able to operate on any unused channel allocated to PA use.

In addition to clearly defined rights, the linchpin of the protected shared-access model is an advanced, highly automated frequency coordinator known as a Spectrum Access System, or SAS. As with LSA/ASA, the CBRS will employ one or more SASs to manage spectrum use in real time using a combination of geolocation and database technologies. This approach builds on one developed for the unlicensed devices that can operate in vacant TV channels (TV white space devices). A communications law blog describes in simple terms how this process would work:

Every device will have to check in with the SAS, report its own location, request permission to transmit and wait to be assigned a specific frequency. The SAS’s job will be to keep everybody off the incumbent spectrum in the exclusion zones, prioritize PA users and assign GAA slots to others.xvi

Although most stakeholders have applauded the FCC’s two-year effort to open the 3.5 GHz band to sharing, some have raised concerns about the complexity of the CBRS plan or questioned whether the technology needed for dynamic sharing is sufficiently mature. (The blog quoted above compared the FCC to an acrobat who is trying to juggle while crossing a tightrope on a unicycle.) Some cellular operators and vendors initially expressed a desire to see the FCC exclude (or delay) the GAA tier, although opposition to a three-tier approach declined over time, and some stakeholders fought for rules that would accommodate the deployment of LTE-U/LAA technology in the third tier. Whatever the concerns with the CBRS, no one disputes that protected shared access (with or without a GAA tier) represents an important new tool for spectrum management—one that can complement exclusive licensing and unlicensed access.

ENDNOTES
i Some advocates of the spectrum commons, or unlicensed, model reject the implication that it is not market-based. In their view, whereas exclusive licensing creates a market in spectrum rights, unlicensed allocation creates a market in the devices and applications that use spectrum.
ii Although some dismiss this sparring as “sibling rivalry,” others recognize it as an important war of ideas. One thought leader sees the intellectual battle between adherents to the property rights and commons models of spectrum governance as the major debate in spectrum policy over the last 15 years. Yochai Benkler, “Open Wireless vs. Licensed Spectrum: Evidence from Market Adoption,” Harvard Journal of Law & Technology, 26 no. 1 (2012). Another participant has observed that this debate lacks the rancor of other telecommunications policy disputes, in part because the two sides listen to one another and take seriously the criticisms of the other side. Adam D. Thierer, “Three Cheers for the FCC Spectrum Task Force Report,” Techknowledge, no. 44, (2002). Available online: http://www.cato.org/publications/techknowledge/three-cheers-fcc-spectrum-task-force-report.
iii William J. Baumol and Dorothy Robyn, Toward an Evolutionary Regime for Spectrum Governance: Licensing or Unrestricted Entry? (Washington, DC: Brookings Institution Press, 2006): 9–14.
iv Ronald H. Coase, “The Federal Communications Commission,” Journal of Law & Economics 2 (1959): 1–40.
v Baumol and Robyn, Toward an Evolutionary Regime, 10–11.
vi “CTIA’s Annual Survey Says US Wireless Providers Handled 3.2 Trillion Megabytes of Data Traffic in 2013 for a 120 Percent Increase Over 2012,” CTIA-The Wireless Association, June 17, 2014. Available online: http://www.ctia.org/resource-library/press-releases/archive/ctia-annual-survey-2013.
vii CTIA-The Wireless Association, Annual Wireless Industry Survey, 2014. Available online: http://www.ctia.org/your-wireless-life/how-wireless-works/annual-wireless-industry-survey.
viii FCC Spectrum Policy Task Force, Report of the Unlicensed Devices and Experimental Licenses Working Group, November 15, 2002. Available online: http://transition.fcc.gov/sptf/files/E&UWGFinalReport.pdf.
These bands were already being used for non-communications purposes and thus were less susceptible to harmful interference. Moreover, they provided sufficient bandwidth to allow for the development of devices that could deliver high data rates.
ix For an inventory of unlicensed spectrum, see Table 2 in Coleman Bazelon, “Licensed or Unlicensed: The Economic Considerations in Incremental Spectrum Allocations,” Communications Magazine, 47 no. 3 (2009): 113.
x Yochai Benkler, “Overcoming Agoraphobia: Building the Commons of the Digitally Networked Environment,” Harvard Journal of Law & Technology, 11 no. 2 (1998): 287–400.
xi Peter Rysavy, “Will LTE in Unlicensed Spectrum Unlock a Vast Store of Mobile Broadband Capacity?” MIMOWorld, June 5, 2014. Available online: http://www.mimoworld.com/?p=2377. See also Preston Marshall, “The View Ahead: Technology Opportunities.” Paper presented at Looking Back to Look Forward: The Next Ten Years of Spectrum Policy Washington, DC, November 2012. Available online: http://www.siliconflatirons.com/documents/conferences/2012.11.13%20Spectrum/PositionPapers/Marshall_TheViewAhead.html.
xii For an analysis of the factors driving the move to smaller-cell architectures, see John Chapin and William Lehr, “Mobile Broadband Growth, Spectrum Scarcity and Sustainable Competition.” Paper presented at the 39th Research Conference on Communication, Information and Internet Policy (2011 TPRC), Arlington, VA, September 2011. Available online: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1992423; and William Lehr and Miguel Oliver, “Small Cells and the Mobile Broadband Ecosystem.” Paper presented at the 25th European Regional Conference of the International Communications Society (Euro ITS2014), Brussels, June 2014. Available online: http://econpapers.repec.org/paper/zbwitse14/101406.htm.
xiii This was one of several actions that the Commission took to refine the LSA/ASA model. See MIT Communications Futures Program, Spectrum Working Group, Toward More Efficient Spectrum Management: New Models for Protected Shared Access, Cambridge, MA 2014. Available online: http://cfp.mit.edu/publications/CFP_Papers/CFP%20Spectrum%20Sharing%20Paper%202014.pdf.
xiv FCC, In the Matter of Amendment of the Commission’s Rules with Regard to Commercial Operations in the 3550-3650 MHz Band, GN Docket No. 12-354, Notice of Proposed Rulemaking and Order, (Rel. December 12, 2012). See also, FCC, In the Matter of Amendment of the Commission’s Rules with Regard to Commercial Operations in the 3550-3650 MHz Band, GN Docket No. 12-354, Further Notice of Proposed Rulemaking (Rel. April 23, 2014).
xv FCC, In the Matter of Amendment of the Commission’s Rules with Regard to Commercial Operations in the 3550-3650 MHz Band, GN Docket No. 12-354, Report and Order and Second Further Notice of Proposed Rulemaking (Rel. April 21, 2015). Available online: http://www.fcc.gov/document/citizens-broadband-radio-service-ro.
xvi “FCC Proposes New Approach to Spectrum Management,” CommLawBlog, April 27, 2014. Available online: http://www.commlawblog.com/2014/04/articles/unlicensed-operations-and-emer/fcc-proposes-new-approach-to-spectrum-management/.
 
 
Title Goes Here
Close [X]