Legal/Technical Architectures of Cyberspace

Spread Spectrum: Regulation in Light of Changing Technologies

Team Members
Stuart Buck
Dedric Carter
Donna Dutcher
Andrew Garcia
Dave Kumar
David Pearah
Andres Rodriguez
Executive summary by David Pearah
Oral presentation by Dave Kumar


Ensuring that all Americans have access to advanced telecommunications services is one of the most important challenges today We must bring real competition to the local loop, not just for telephony and low speed data, but for broadband applications as well. Wireless should be a key driver of this revolution.
- William Kennard, Chairman of Federal Communications Commission
- Speech to Personal Communication Industry of America, 23 Sep 1998
Though electronic communications is just over 100 years old, it has left an indelible mark upon the global society and economy. Telecommunications has paved the road to an information economy, and has challenged the limitations of space and time, whether conducting electronic commerce or simply a conversation across the world. While electronic communication has long undergirded the fabric of the United States, the past few years have witnessed an increase in the nation's awareness of its impact and importance. If the popular press is any indicator, telecommunications will firmly remain within the public discourse for the foreseeable future.

While this public dialogue generally centers on technological change, there is increasing scrutiny of the role of Congress and its agencies in the development of telecommunications. The seemingly perpetual mergers of communications giants, coupled with the oft-heralded failure of the Telecommunications Act of 1996 to provide meaningful residential broadband competition, have created concern over the administration and ownership of this vital service. The threat of consolidation of the means of advanced communications legitimates reexamination of traditional regulation and begs exploration of alternatives.

This paper participates in this critical investigation by looking at a particular instance of communications: spread spectrum packet radio networks. The reason for this focus is two-fold. First, though the technical and specific title would suggest otherwise, the implications of this application are far-reaching. If regulation is the "practice" of timeless law in a certain context, then the very constitutional legitimacy of spectrum regulation must be questioned in light of technological advance. Second, in contrast to cable and ILEC competition that requires the construction of expensive wireline links, wireless communications is often viewed as the last/best hope of supplying meaningful broadband competition (and by extension, business and educational opportunities) to residential markets. Since at-cost leasing (i.e. unbundling) of incumbent communications equipment and services has met with limited success, regulators are under increasing pressure to review the effectiveness of their regulatory practices and to encourage alternative forms of communication.

The organization of the paper is as follows: Section 2 covers the fundamentals of wireless spectrum and how spread spectrum facilitates communication without explicit coordination. Section 3 provides a critical history of spectrum regulation, tracing the evolution from anarchy, to lottery in the public interest, and ultimately to auctions. Legal, regulatory, and technological recommendations are the contributions of Section 4, followed by concluding thoughts.

SECTION 2: SPREAD SPECTRUM FUNDAMENTALS

Section 2.1: What is Spectrum?

While the technology focus of this paper is wireless communications, "spectrum" is a general term to encompass both the spatial and temporal properties of any medium, including fiber optics and the ambient air. Restricting the term to the more common application of communications, the term "spectrum" represents the temporal and spatial opportunities to transmit information. A multilane highway provides a natural and useful analog: the opportunity of a car to move along a highway is represented by the empty spaces both within and between lanes. By associating "frequency" with the lane choice and "time" with the distance along the highway, the spatial and temporal division of the highway becomes clear.

In the same way, the medium of wireless communications (i.e. "ether") can be divided both in frequency and time. In order to transmit information, it must be encoded to transmit at certain frequencies and/or within specific time slots. FM radios provide a common example of frequency division multiplexing: the listener adjusts the knob to accept a music station transmitting at a certain frequency and to reject all other signals. Computer terminals on a mainframe network illustrate time division multiplexing: the central computer divides time into slots, and dedicates its total processing power to each user for 1 slot of time in turn. The next generation of cellular phones (CDMA) demonstrates the simultaneous application of both time and frequency division multiplexing.

The highway analogy also provides insight into the finite nature of spectrum. At a given state of technology, the amount of spectrum that is available for communication is necessarily limited. Just as a highway can grind to a halt both within and between lanes, information can be halted due to a lack of opportunities to transmit. As technology progresses, data can be sent at increasingly higher frequencies and with finer temporal granularity, thereby bringing more spectrum into productive use. But at any given stage of development, the amount of "usable" spectrum is indeed finite.

Another aspect of spectrum is reuse. Consider the case of two radio stations separated by a large distance. If the two stations transmit on the exact same frequency and limit the power at which they transmit, the two signals will not interfere. Since signal power drops off as the inverse square of the distance away from the transmitter, a communications system can be devised to judiciously place transmitters in slightly overlapping cells, almost doubling the capacity of the system in the single transmitter case. The wireline equivalent of cell reuse is to simply lay down another line between the communications points, effectively doubling the capacity of the circuit.

Section 2.2: Spread Spectrum

Spread spectrum is a class of modulation techniques developed over the past 50 years. To qualify as a spread spectrum signal, 1) the transmitted signal bandwidth is greater than the baseband (i.e. minimal) information bandwidth, and 2) some function other than the information being transmitted is employed to determine the resultant transmitted bandwidth.

Most commercial spread spectrum systems transmit an RF signal bandwidth as wide as 20 to 254 times the bandwidth of the information being sent, while some have gone as high as 1000 times. Common spread spectrum systems are of the "direct sequence" or "frequency hopping" type, or else some hybrid of these two.

Direct sequence spread spectrum systems derive their name from their high-speed pseudo noise code sequence. This sequence directly modulates the carrier signal, thereby directly setting the transmitted RF bandwidth. This modulation produces a signal centered at the carrier frequency. In contrast to direst sequence, the spectrum for frequency hopping systems is generated by "hopping" from frequency to frequency over a wide band. The specific order in which frequencies are occupied is a function of a pseudo noise code sequence, and the rate of hopping from one frequency to another is a function of the information rate. While the transmitted spectrum of a frequency hopping signal is quite different from that of a direct sequence system, it is sufficient to note that the data is spread out over a signal band larger than is necessary to carry it. In both cases, the resultant signal appears noise-like and the receiver utilizes a similar technique to recover the original signal.

The noise-like character of the transmitted signal drastically reduces the probability of signal detection and interception, a fact that the U.S. military exploits to ensure secure communication. But strong encryption and spoofing countermeasures can be added (perhaps at great cost) to existing narrowband communications schemes. The property of interest is spread spectrum's ability to provide point-to-point communication without explicit coordination of the speakers. A crude analogy can be made to the CB radios that truckers often employ: the speaker keeps switching the channel until a free spot is open. Spread spectrum's more sophisticated hopping sequence "spreads" the speaker's message over various channels at different points in time. This pseudo-random hopping behavior unseats the long-held assumption that signals from two more speakers may not overlap in time and space in order for communication to occur. To the contrary, all spread spectrum systems have a threshold or tolerance level of interference below which useful communication continues unimpeded.

As the tutorial above demonstrates, spread spectrum technology challenges one of the fundamental justifications for governmental administration of spectrum: scarcity due to signal collision by different speakers. Even though there is a limit beyond which this technology cannot recover interfering signals, regulation is a matter of degree in a technology context--there is sufficient evidence to warrant a reexamination of spectrum management, both in the past and present.

SECTION 3: SPECTRUM MANAGEMENT

While the Telecommunications Act of 1996 marks a significant milestone in the legislation of wireline communications, wireless law has emerged relatively unchanged from its inception in the Radio Act of 1927. However, the application of the law (i.e. wireless policies and regulations) has had a more varied and colorful history.

Section 3.1: Electromagnetic Anarchy to Lottery

The earliest widespread usage of radio technology provided ship-to-shore and ship-to-ship communications under the aegis of the Navy. Almost all radio communications regulated by the Radio Act of 1912 was wireless Morse code transmissions. It was not until after World War I that private enterprise began to take hold of the industry while the Navy decreased its involvement. In a plan to fuel the radio market in the 1920's, a number of equipment providers started broadcasting stations to encourage adoption of the product. Since radios of that time employed a simple frequency division multiplexing scheme, the lack of central coordination coupled with a growing subscriber base resulted in severe signal interference between stations.

In order to preserve the radio industry, the Department of Commerce provisionally issued licenses to stations and imposed restrictions on frequency, power and hours of operation. When the courts halted this type of regulation, chaos reigned as stations constantly hunted around the spectrum for better broadcasting frequencies. In response to the court decision, the legislative branch quickly approved the Radio Act of 1927. Instead of the Department of Commerce, the regulatory fiat was conferred upon the newly minted Federal Radio Commission, only to be replaced soon after by the Federal Communications Commission (FCC) in 1934. The FCC was given even broader authority to protect radio communications as a public good, under the premise that spectrum was a scarce commodity that required central coordination. In order to serve the public interest, the FCC used first a review process and then a lottery system to distribute monopoly licenses within frequency bands.

Section 3.2: Today's FCC

Title 47 of the US Code of Federal Regulations enumerates the responsibilities vested in the FCC, both in allocation and licensing. While a large part of the spectrum is earmarked for government use, the FCC holds jurisdiction over the rest of spectrum. The FCC is charged with the following allocation duties: 1) dividing the available spectrum into channels, 2) mapping specific communications uses to sets of channels, and 3) establishing transmission rules (e.g. power) for each channel set. While the FCC's position is to remain competitively neutral in making these decisions, the process is invariably politicized and volatile. For example, determining the amount of spectrum for PCS services will in turn shape the number of competitors within a given market and the nature of the service (e.g. beepers v. phones). These decisions carry great weight in supporting certain technologies while creating large barriers to entry for other services.

In addition to allocation, the FCC is charged with issuing licenses that further the public interest, convenience and necessity. Licenses are valid for a maximum of 8 years--spectrum cannot be owned but rather leased by a private interest. When the licenses expire, the agency can review and reallocate the size and purpose of the bands (e.g. HDTV initiative to reclaim spectrum). Specific licenses encapsulate the allocation rules above, so even if a license is transferred or sold, it may not be used for any purpose other than that intended by the FCC. The agency can chose to set aside frequency bands for unlicensed (e.g. 3 U-NII bands) and experimental operation, but it always retains the right to ordain transmission rules and to police users.

The FCC slowly made the transition from lotteries to auctions to streamline its licensing responsibilities. After congressional passage of the Omnibus Budget Reconciliation Act of 1993, the FCC began to license monopoly rights to radio spectrum via competitive bidding. Pursuant to this mandate, the FCC created the Wireless Telecommunications Bureau (WTB) on 1 December 1994. As of 22 October 1998, the WTB has conducted 18 separate spectrum auctions, with a total of 6,801 licenses awarded and governmental receipts of $22,903,134,860. Armed with continued auction authority conferred by the Balanced Budget Act of 1997, the FCC recently announced its intentions to expand auctions to "commercial radio and analog television stations."

Under the assumption that spectrum must be narrowly and strictly regulated, auctions have been heralded as fair and efficient for the end users, as well as fiscally responsible for the government. The most common argument in support of this practice is the claim that selling spectrum rights will lead to the greatest economic efficiency--putting spectrum into the hands of those who value it most highly. This is difficult to verify empirically since the definition of "value" used here is circular: the highest-value user is defined as whoever happens to win an auction, and then the auction is praised for its power of discovering this user's value. There may be many combinations of competitive providers that could ultimately produce greater social and economic value, but the nature of the winner-take-all auction predicts that either 1) a monopoly firm will win or 2) a consortium of firms will win but then fail to compete against the other members. For example, if the socially optimal system is 3 PCS providers utilizing a spread spectrum system instead of designating a single winner, auctions (as currently implemented) will always fail to achieve the desired result. In addition, auctions create huge barriers to entry for minority service providers, and the regulatory Band-Aid of providing "free" auction credits to these businesses is hardly optimal and infrequently successful.

Other arguments in favor in auctions, e.g. quick implementation and promotion of technological innovation, may indeed prove true in light of previous management schemes, but these are not unique to the auction system. In the next section, a new model of spectrum management, one that accounts for the new technology context created by spread spectrum, is presented as an alternative to the traditional auction system.

SECTION 4: LEGAL & TECHNOLOGICAL RECOMMENDATIONS

Section 4.1: Freedom of Speech

The First Amendment right to free speech provides a compelling reason for retooling the way the FCC regulates spectrum; indeed, it was the threat of losing this right that invited this strict regulation in the first place. Free speech requires that any intervention must be narrowly tailored to further a substantial public interest. Where regulation once required centralized allocation to facilitate a spectrum-based communications marketplace, spread spectrum technology has matured sufficiently to enable speech without explicit coordination.

There is certainly a limit to spread spectrum's ability to negotiate information transfer in a completely unregulated environment--it is not an engineering panacea that dispenses with all regulation. The central issue under examination is the nature and timing of the government's role in administering the spectrum space. Since this technology enables uncoordinated communication in a fairly robust way, minimal interference with the freedom of speech demands that there be a compelling public interest to justify the regulation of ether. The government should intercede only when it is necessary to preserve speech. By using the hammer of regulation over all wireless communication, it violates the First Amendment Rights of those who would have used the technology. To reiterate: the government's management of spectrum violates the Constitutional rights of those for whom scarcity does not exist.

Certainly, in high-density areas where information exchange is intense, the government has a positive and important role to play, e.g. coordinating cell sizes and enforcing interconnection between providers. In addition, it may assume a role of leadership in defining electronic standards to ensure interoperability between equipment providers. However, the government must not arbitrarily assume that scarcity exists and use that reasoning to restrict the speech of individual citizens.

The constitutionality of a regime that awards exclusive speech rights to a select few is tenuous in light of a technology that does not require dedicated spectrum bands. A reviewing court would likely strike down the present allocation system as too great an interference with the right to free speech: it deprives potential speakers of a voice in a manner that is not narrowly tailored to meet communication interests.

A strong analogy can be drawn to the regulation of highways. The government has a positive role in establishing maximum speed limits, requiring safety devices such as seat belts and headlights, enforcing maximum car dimensions so that everyone can share the road, and polices the entire system to remove those who abuse it. However, the government certainly refrains from creating regional car monopolies and most assuredly does not tell citizens when and where to drive. Recognizing the power of standardization to encourage freedom and innovation is a delicate balancing process, but resorting to wholesale control of every facet of the driving process would clearly be a violation of freedom.

In the same manner, while wireless communication still merits governmental coordination in the form of standardization and interconnection, unilateral control of wireless speech clearly goes too far, especially in light of spread spectrum techniques.

Section 4.2: Spectrum as the Commons

Rather than a parcel of land that an individual entity owns for a period of time, spectrum should be viewed as a public good--a commons. In this regime, the commons is for public use provided that citizens abide by minimally invasive guidelines. A regulatory body, or an authority acting under its aegis, could govern issues such as cell reuse, code standardization, signal power, and bandwidth usage. Policing would occur in much the same way laws of the commons function in real space. What is necessary is creation of the electromagnetic equivalent of nuisance laws, replete with reasonable time, place, and manner restrictions that are permissible under the First Amendment.

Instead of imposing the auction paradigm on all situations, market mechanisms can be tailored appropriately for the local demand for information transmission. For a rural schoolhouse that requires broadband Internet access, but cannot afford to pay the many hundreds (if not thousands) of dollars per month for a wireline solution, there need not be any explicit market mechanism. In this case, since there is presumably no contention for spectrum resources, the only necessary costs should be the capital equipment costs.

A middle class neighborhood provides a good example of a moderately dense area with a fair amount of communications needs. The municipality may assist in organizing a town coalition to establish a network of radios and repeaters to provide basic data service to its residents. If demand increases, a central tower, operated by either the town or an outside entity, may provide a higher grade of service and a larger link to the Internet. So long as demand does not exceed the available useful spectrum, there need not be centralized control of who may transmit.

In a dense metropolitan area with high demand for data, there would need to be explicit coordination of locating cell structures and perhaps mandatory interconnection of cell networks. A real-time spot market for spectrum can be created to provide varying levels of service to users. Whatever the case, the government need not declare scarcity prematurely and then carve the spectrum into oligopolies. If and when scarcity rears it ugly head, only then does the government have a constitutional mandate to introduce market (or non-market) mechanisms to address the shortage.

Unlike the current system spectrum auction and oligopoly operation, these solutions encourage diversity in both applications and speakers. By reducing the barrier to becoming a "station," minority broadcasters need not rely on alliances and subsidies to operate a business in the spectrum space. By providing a robust bit carriage service rather than a specific application (e.g. pagers), diversity of applications is motivated much as it is on the Internet (e.g. WWW, telnet, ftp). While it may not be technologically possible to ensure strict quality of service guarantees over the open access wireless networks proposed here, this does not imply that the service has no value. Just as the Internet does not guarantee timely or even eventual delivery of packets, users may elect the service if it suits their needs. If a user requires a service that is not provided by the commons, she is certainly free to utilize other private services, so the lack of specific transmission guarantees should not be an impediment to adoption.

CONCLUSIONS

There is no doubt that the spread spectrum frameworks suggested above challenge the traditional assumptions under which spectrum is managed, but they also affront the very notions of carrier and user. Under traditional monopoly/oligopoly regulation regimes, both the legal and regulatory codes make great efforts to differentiate the rights and responsibilities of these two entities. But if we allow users to become their own carriers, should they be subject to the same reporting requirements, as are traditional cellular providers? Must an ad hoc network of spread spectrum radio users pay into a universal service fund that supports wireline telephony in their own town? Many of the existing management principles cannot account for the diversity of both speakers and listeners, much as the Internet frustrates the monikers of local- and long-distance.

While the transition to a minimally interfering management regime will no doubt prove to be a Herculean task, both technically and politically, it is the government's duty to support and enforce the precepts set forth in the Constitution. Even if it is currently infeasible to transition the traditional wireless players to a spread spectrum system, any available and future spectrum must be treated as a public good. The regulatory powers that be must be vigilant in examining the assumptions upon which they suspend the freedom of speech. Spread spectrum technology has forever changed the technology context: strict, central administration of the spectrum space is not necessary to ensure communications, so it must not be treated like private property that can be leased and re-leased at the whim of a few large industry players.


Return to conference page


Send comments about this site to 6805-webmaster@martigny.ai.mit.edu.


Last modified: December 3 1998, 9:35 AM