MARCONI CENTENNIAL SYMPOSIUM Bologna, Italy June 23, 1995 Is the UHF Frequency Shortage a Self Made Problem? Paul Baran Atherton, California, USA THE FIRST HUNDRED YEARS Today on this occasion of the Marconi Centennial, we celebrate the accomplishments of the first hundred years of radio. The history of radio has been one of rapid and continuing progress. No letup is in sight. Each generation sees major advances over the previous art. Yet throughout the history of radio, scarcity of spectrum has been a fact of life. Lack of spectrum limits progress in creating new applications. The constant shortage of spectrum space is not a new issue. One of the very first questions asked of young Marconi about his nascent technology was whether it would ever be possible to operate more than one transmitter at a time. Marconi's key British patent No. 7,777 was a milestone as it taught the use of resonant tuning to permit multiple transmitters to simultaneously occupy the radio spectrum. New demand is like a constant vacuum, sucking up frequencies as they become available. Even today, with over 30,000 times more spectrum at our disposal than in Marconi's day, entrepreneurs wishing to implement new services encounter the same perpetual shortage of frequencies. REGULATION Although it's been called ether, the radio spectrum really is a lot like money. No matter how much some people seem to have, they always want more. Enough never seems to be enough. Given the limited spectrum available combined with the growing demands of potential users, it became necessary early in the game to devise some means of rationing the spectrum resource. National and international regulatory structures evolved over time primarily to restrict access to the spectrum only for specific, allocated uses which were then further limited only to chosen institutional entities - called licensees. Today, we have new technologies that could potentially ameliorate this perennial shortage. But such technologies cannot be fully utilized because the very regulatory system initially set up to address the frequency shortage of the past stands in the way of the present and the future. History has shown that there are very few mechanisms as effective at maintaining the status quo as a set of institutionalized regulations. Once set in regulatory concrete, reconsideration of the basic underlying assumptions is very difficult. While it will be an uphill fight to re-examine the basic underlying assumptions of any law or administrative rule, it is clearly not impossible. It will just take longer than if not so well institutionalized. So on this day, as we celebrate the first one hundred years of radio, let us take this occasion to review some of our basic assumptions about spectrum utilization and to consider a possible alternative approach to frequency regulation in the future. My words this afternoon focus on the UHF spectrum, 300 to 3,000 Megahertz, that most desired part of the radio spectrum for communications with high data rate for local area data devices. These are the frequencies preferred by designers seeking to use low cost electronic products to deliver new services. Today, the UHF band is the carrier of the bulk of terrestrial radio services -- cellular telephony, broadcast television, cordless telephones, etc. And, it is used for low altitude satellites as well. A PARADOX To suggest that there really is no fundamental reason for a shortage of UHF spectrum is to violate the common wisdom. Tune a spectrum analyzer across a band of UHF frequencies and you encounter a few strong signals. Most of the band at any instant is primarily silence, or a background of weaker signals. The spectrum analyzer connected to an antenna reveals that much of the radio band is empty much of the time! This unused spectrum might be available for transmission if we could take measurements and know exactly when and where to send the signal. In part, the frequency shortage is caused by thinking solely in terms of dumb transmitters and dumb receivers. With today's smart electronics, even occupied frequencies could potentially be used. DIGITAL VERSUS ANALOG To the modern communications engineer, a lack of strong signals anywhere, no matter how distributed, represents a theoretically unused capacity that is available to be utilized with the proper signal processing. With advanced signal processing techniques, transmission of signals on top of undesired signals received at lower levels represents a potential source of usable transmission capacity. There is a caveat here. We are assuming digital signals that are able to operate with lower signal to noise ratios than analog signals. That means if the desired signal is but slightly stronger than an interfering signal, it can theoretically be received without error. This game doesn't work with old fashioned analog modulated signals, such as analog broadcast TV signals where even weak interference 40 dB below the picture is visible. 40 dB is a power ratio of 10,000 to 1. That means if an interfering signal is 1/10,000 as strong as the analog TV signal it will be visible in the received TV image. Compare this situation relative to the case of a digitally modulated signal able to operate at a 20 dB signal to noise ratio. 20dB is a power ratio of 1 to 100, or a tolerance 100 times as great as in the analog TV case. With the addition of error correction codes, some digital systems can operate at a 10 dB level or a noise tolerance 1000 times as great as our analog TV example. How much new usable capacity could we gain by using digital transmission? Think in terms of a curve of energy versus frequency at the receiver. The potentially available bandwidth can be visualized by inverting this received energy versus frequency curve and then adding a second curve above the first curve separated by an amount equal to the required signal to noise ratio. This new curve suggests the amount of potential spectrum actually available for reuse using the improved modulation. In practice, the benefits of using digital transmission for voice signals is not as great as for our TV example. Analog voice is able to operate with lower signal to noise ratios than television. Nevertheless, the savings in the evolution from analog to digital cellular/PCS voice systems is worthwhile. In reality, the major spectrum hog is analog broadcast TV transmission. In the US and to an extent in other countries a spectrum analyzer will find much of the allocated VHF and UHF TV spectrum unused, even in big cities. The UHF television band is punctured with vast empty holes called "taboo channels". These channels are left unoccupied because of the frequency selectivity limitation of early era television receivers. Today we know how to build far better receivers than when this early rule was adopted and when those frequencies were set aside. We should never forget that any transmission capacity not used is wasted forever, like water over the dam. And, there has been water pouring here for many, many years, even during an endless spectrum drought. LINK VERSUS SYSTEM THINKING We have briefly considered digital's greater tolerance to interference and how this can be translated into better spectrum usage. But, this direction offers a relatively minor improvement as compared to other possibilities. To better capture these additional potential savings, it is necessary to think in terms of networked systems rather than single links. Our understanding of the issues is most highly developed when considering how to make best use of a single communications channel or a single link. For example, in this symposium Joachim Hagenauer's paper, Approaching Shannon's Limit on Radio Channels focuses on channel optimization. A second domain of interest is in making best use of a single entity cooperatively sharing a limited spectrum which is considered by Andrew Viterbi in his paper, Universal Wireless Communication. Here he addresses the case of making best use of a band of frequency by a homogeneous assemblage of units - such as cellular telephones - operating with a common standard. The third domain and the subject area that is most ripe for advancement is the focus of this paper -- learning how to optimize the overall interests of a multiplicity of competing heterogeneous users, each with different requirements, and all sharing a common block of shared frequencies. The challenge is how to make best use of a common shared spectrum to handle disparate users, modulation and applications in a world of different system technologies, different system owners and different needs and objectives. The argument will be made that instead of today's regimented channel by channel, highly centralized form of regulation, an alternative approach requiring only a minimal measure of cooperation would work to the maximum benefit of all. A LINK IS NOT A SYSTEM When wireless first started, a system comprised a transmitter and a receiver. Just as the telephone has evolved from being a pair of instruments connected by a pair of wires to a switching network, so too has radio moved from the transmitter and the receiver pair to becoming part of a larger switched system. Communications networks are designed by choosing and joining together a combination of different media links and switches, as no single communications medium is ideal in all situations. If the link requirement is for long distance transmission, then optical fiber may be used. If the requirement is to address many users located hundreds of meters apart form one another, then coaxial cable or twisted pair may be the preferred medium, depending on the data rates for that part of the network. One example of such a composite network is the cellular telephone network, composed in part by radio links integrally attached to the switched telephone network. Such networks are owned by a single entity and tend to be reasonably well optimized, with the economic factors considered as a whole by the network designers. As we build more of these kind of networks in the future we are likely to find that wireless will increasingly becoming the preferred medium, for the tails of the network - allowing "tetherless" operation. This composite arrangement provides freedom and flexibility to the user and it combines access to the more cost effective longer distance transmission media. RANGE REDUCTION When we combine the radio tails with wired portions of the network, we face a tradeoff as to the amount of each media is used. In the UHF band the number of geographically dispersed users that can be simultaneously accommodated by a fixed spectrum varies as the inverse square of the transmission distance. Cut the range in half, and the number of users that can be supported is doubled. Cut the range by a factor of ten, and 100 times as many users can be served. Reduce the power further, and then essentially any number of users can be fit into the exact same spectrum presently tied up in supporting a few longer distance, higher power users. In other words, a mixture of terrestrial links plus shorter range radio links has the effect of increasing by orders of magnitude the usable frequency spectrum. We speak of inverse square ranges. That is valid for free space attenuation. When it comes to the real world, the payoff is even more dramatic. For example, radio signal attenuation within concrete office buildings is closer to the inverse 4'th power of the transmitted power. Given the high attenuation in those environements, increasing the radiated power doesn't buy much in the way of range improvement anyway. By authorizing high power to support a few radio users to reach slightly longer distances, we deprive ourselves of the opportunity to better serve the many. Automatic power reduction increases the number that can use a shared spectrum. COMPOSITE PATH How realistic is it to reduce the range of transmission for the relatively few to allow a greater number to benefit? Consider today's millions of short range cordless telephones all sharing a minuscule slice of the radio spectrum, while a small number of licensed users occupy most of the spectrum. Most of these could be served by shorter range radios plus a telephone of fiber line to provide the longer distances sought. While the resulting path is not all wireless, neither is today's cellular systems. The advantage of tetherless operation is retained for the user's convenience. There is very little "give up." In the US, for example, and, increasingly in other countries, there is an underutilized transmission capacity in already installed TV coaxial cable and the telephone twisted pair plant. Assuming a move to this direction, a vast communications capability to homes and businesses can be created to allow the support of a far greater number of users with greater bandwidths than feasible today. MOVING RADIO BASED TV TO CABLE We could significantly increase the available UHF bandwidth by giving each TV viewer "free" access to a TV cable to receive the present few off-the-air signals that they now receive. Let's look at the economics. In the United States, TV cable passes about 94% of the households, with 63% already connected. How much would it cost to connect everyone to cable and recycle the released bandwidth at a cost? How Much Will it Cost? Since TV cable systems are laid out without knowing which houses will take TV service or not, taps to serve each potential subscriber are in place. No additional power is required. The incremental cost of running a drop cable to each house is on the order of about $40 per house. How Much Would be Saved? Almost 500 MHz of spectrum is presently assigned for over-the-air TV transmission. In the US the Federal Communications Commission recently raised about $8 Billion selling access to about 70 MHz of UHF spectrum for Personal Communications System services. This is about $80 per US household, or about $1.14 per household per Megahertz of bandwidth. And this cost only covers the cost of the license paid to the US Government - before any actual equipment is deployed! If we assume that the TV band occupies about 480 MHz (80 channels) of spectrum, the value of this TV spectrum asset if sold on a comparable basis, would be worth about $547 per household, or about 13.7 times the installation cost of the new drop cables. What Does This Mean to the Cable Operator? The cable operator would, of course, lose some revenue if each house were to be given free off-the-air signals. But, the number of cable subscribers who presently pay for off the air signals alone is small, and the received revenue is not large. The loss of revenue to the cable operator is minor compared to the potential revenue for higher value services made possible when the cable enter the houses of that one third of the population not presently reached by the TV cable. And, any short fall might be covered by the alternative revenues received from freeing up the UHF spectrum. What Does This Mean to the Broadcast Station? The TV broadcast station would now be able to reach all their present viewers and no longer have to pay for the TV transmitter costs. "Owning" the TV license places the broadcast station in a position to make far more money leasing frequencies than operating a TV transmitting station. Of course, the issue of who really owns the spectrum is an interesting issue of public policy. We shall ignore this issue other than to note, that if economics are considered most broadly, TV broadcasting is probably not the most economical use of the spectrum. As a practical matter, if this direction was to evolve it would likely occur in an incremental manner, possibly with TV broadcasters cutting deals with cable companies to allow freeing the radio channels. EVOLUTION TO DIGITAL With the movement of TV to cable, digital modulation is likely to be increasingly used. Digital modulation is already being used in early trials on TV cable systems. Digital in lieu of the present analog modulation allows ten times as many TV signals to be sent over an existing TV cable. For example, the TV cable currently carrying 50 analog channels would be able to carry 500 TV channels. SMART TRANSMITTERS We don't have to wait for the eventual transfer of the UHF TV spectrum to cable. The existing spectrum can be more efficiently used by the use of smart receivers and transmitters. Inexpensive microcontrollers would be used that first listen and then automatically choose preferred frequencies to avoid other signals in the band. It is really a matter of being a good neighbor. The smart transmitter reduces its power level to that needed to produce an error free signal and no more. A pristine pure slice of spectrum to have error-free performance is not required when using digital modulation. Digital logic on a chip implements error correcting codes to convert a small amount of redundancy in transmission enabling even highly corrupted signals to be cleaned up to emerge error free. SPREAD SPECTRUM One particularly interesting variant of digital - spread spectrum modulation can allow more users to share a common band of channels. But, there is a regulatory lag in encouraging the fullest use of such technology as spread spectrum seems to require more spectrum space. The idea of signals taking more bandwidth is at variance with the mind set of government regulators whose objective has historically been to minimize the occupied bandwidth. And, this takes us to our next topic, Regulation. REGULATORY HISTORY Given the history of the shortage of spectrum leading to the necessity of rationing, it is understandable how national and international regulatory structures evolved, concerned in major measure with doling out bandwidth. The assumption of shortages is so institutionalized into regulatory policy that the basic assumptions that got us here rarely ever get re-examined. And, when they are, changes tend to occur at glacial speeds. We have a wide range of sophisticated but under-utilized technology with which to address this problem but there is a roadblock as our institutionalized regulatory structure with its implicit assumptions that spectrum is a scarce commodity like real estate, leading to a zero-sum game. While this view of the spectrum may have been valid once upon a time, it is less so today, and will not likely be true tomorrow. But, the rules of the regulatory game are set by governments, while the issues are primarily technical. Government agencies tend to be staffed by lawyers who view a frequency as a unique property right If I owned a frequency, then you can't use my frequency. It's mine, exclusively mine. Yet, communications engineers know that statistical averaging of larger blocks of frequencies can allow for better usage. That is what cellular radio is all about. There was a regulatory delay in the allowance of cellular telephones in the US for well over a decade. In fairness, newer thinking is increasingly being incorporated in the regulatory decisions. But, from the point of view of a technologist, the process is agonizingly slow and in need of rationalization. ALTERNATIVE DIRECTIONS Given the technological options described above, the assumption that there will always be a shortage of UHF frequencies deserves reconsideration. If our present regulatory approach is lacking, how can we do better? It is my belief that public policy might be better served if we moved to an environment of near zero regulation. In such an environment anyone and everyone would be allowed to use the spectrum, without the barriers to entry that keep out the true innovators. Of course, there will be some minimal rules necessary, such as maximum allowable transmitted power and power densities. The micro-managed regulatory approach of today, such as who can use any single frequency is neither necessary nor desirable. If the hypothesis is correct that there is a potential for a limited amount of spectrum to carry all the traffic imaginable (assuming that the power and the range of the transmitters is limited), then the purpose of regulation would no longer be primarily keeping potential users away from the spectrum. CHAOS? Would this laissez faire form of regulation lead to chaos? Possibly, but most likely not. Consider the many millions of cordless telephones, burglar alarms, wireless house controllers and other appliances now operating within a minuscule portion of the spectrum and with limited interference to one another. These early units are very low power 'dumb devices' compared to equipment being developed, able to change their frequencies and minimize radiated power to better avoid interference to themselves and to others. Of course that means that there will have to be enough frequency spectrum set aside to do so. But, once having done so, we would have created a communications environment able to handle orders of magnitude more communications than today. REGULATION FOR THE FUTURE --THE INTERNET MODEL The Internet provides an instructive model for the future of telecommunications regulations. The Internet allows worldwide communications at a far lower cost than any alternative; serving data users inexpensively, and opening access to the world's information to a greater number of people than ever initially imagined. In the Internet, there is no central node, and only a minimal centralized management structure, limited to a few housekeeping functions such as standards setting. Local decisions essentially control the network. The independent pieces of the network operate in a coordinated manner with a minimum of restrictions. This lack of a limiting centralized structure has permitted the Internet to be responsive to a very large unregulated constituency and allowing explosive growth and with increasing usefulness to its users. Probably the closest parallel structure to the Internet is the free market economy. We know that works. Will it work for regulating the radio spectrum? The Internet is an organization of users sharing a common resource, as appropriate to the sharing of a common band of frequencies by all comers. The Internet model for regulation would be similar to the data network in which each user follows a simple set of commonly observed rules. Which frequency to use and when, or which form of modulation to use would be left to each user. The Internet model has many of the characteristics of a desired communications regulatory approach for the future. Such a direction does require a big evolution in the thinking of the current communications regulatory agencies. The present regulatory mentality tends to think in terms of a centralized control structure, altogether too reminiscent of the old Soviet economy. As we know today, that particular form of centralized system didn't work all that well in practice and, in fact, ultimately broke down. Emphasis with that structure was on limiting distribution, rather than on maximizing the creation of goods and services. Some say that this old highly centralized model of economic control remains alive and well today -- not in Moscow but, rather, within our own radio regulatory agencies.