Law in the Internet Society

Cognitive Radio in the Television Bands

Julian Dunn


Nearing the end of his tenure, Commissioner Michael Copps reflected in late 2008 that one of the great lessons he learned while at the FCC was the power of technology to turn scarcity into abundance. (FCC Docket 08-260) This comment illuminates the core challenge of spectrum policy in America, promoting the highest uses of public resource with minimal offense to the constitutional commitments to equality and free speech. Advancements in spectrum technology have enabled the transition from analog to digital, freeing up spectrum that raised $19B for the Treasury at auction. But Copps' statement was made in reference to cognitive radio, which the FCC tentatively approved for unlicensed use in the vacant TV channels.

On its face, the decision seemed not to be of much consequence; Television Band Devices (TVBDs) are only allowed on a “secondary basis” in the TV Bands, essentially operating in the shadow of broadcasters. But advocates of cognitive radio claimed the acceptance of cognitive radio by the FCC represents a fundamental change in the way information will be sent and received over the air. Cognitive radio allows a device to alter its transmissions based on its surroundings in order to make more efficient use of available frequencies, squeezing value out of otherwise fallow frequencies.

Unfortunately, although the FCC voted unanimously to approve the devices, broadcasters were able to put a thumb on the scale, skewing the ground rules for the TV Bands to favor incumbent interests. Existing uses in the TV bands were able to secure concessions from the FCC and white space advocates, without sufficient justification as to why such a move is in the public interest. While TVBDs are asked to bend over backwards just to be permitted on the air, broadcasters were not required to make any accommodation for the entrant to optimize traffic. TVBDs are essentially treated as a nuisance, even though they are capable of supporting more diverse applications and a greater number of users, using less bandwidth.

  • "More diverse applications ... using less bandwidth" than what?

Placing the burden of coordinating uses entirely on the shoulders of TVBDs increases the barriers to entry, slowing – or potentially halting – the diversification of media and the closing of the digital divide. In most major markets white space service providers may find there is too little spectrum remaining to develop a viable product, let alone a service that could challenge cellular and wireline incumbents. Rural deployments have the advantage of more empty airspace, which will provide a deeper pool of bandwidth to share, but Wireless ISPs (WISPs) in these areas face additional obstacles that may keep them from viability. With these disadvantages it will be exceedingly difficult to put pressure on incumbents to innovate and lower their prices. As a result, the anticompetitive effects of spectrum oligopolies – oligopolies supported via broadcast licenses – will continue to suppress free speech and media diversity.

At present there is too much uncertainty to say just how much these limitations will hinder the development of white space networks and cognitive radio. The two sides of the debate thus far have largely crafted to sway public opinion instead of address the key issues: Broadcasters harp on the ‘harmful interference’ that TVBDs will cause, while white space advocates have tried to create the impression TVBDs will bring of ‘universal free broadband.’ There are still a number of internal decisions and external trends that will determine how this technology can be deployed, and how disruptive it will be in the marketplace.

  • *Been* largely crafted?

In any case, value will be created as this technology matures. But such gains must be evaluated against alternate courses of action that might produce more immediate and meaningful benefits. The Second Report and Order, which outlines device specifications and transmission protocols for TVBDs, reflects an antiquated view of the FCC’s role as a “traffic cop,” rooted in the Communications Act of 1934. While Red Lion still stands, the facts underlying its rationale no longer hold. While broadcast remains an important form of local and national media, is it no longer the most powerful or cost-effective way to receive information. Broadcasters were recently given 'beachfront property' spectrum in their DTV licenses, but they are simply not putting to the highest use. Given the numerical scarcity of broadcast channels in a given market, each license may confer market power on the licensee. The FCC’s continued preference for one-way broadcast, despite the fact that allocational scarcity no longer requires it, causes harm to consumer welfare by sustaining the high level of concentration and lack of diversity in media. Internet access provided over this spectrum would allow for more diverse and meaningful links to society.

  • This is the crux of the argument. I appreciate that to you and me it seems to need no explanation or expansion, but where the usual mechanism of argument is to show there's no reason why we can't have both, sharing broadcast spectrum with bidirectional commodity wireless networking appears to create a more direct conflict situation between broadcasting activities and real networking, which means that a clearer and more compelling explanation of the greater marginal value of the Internet needs to be given.

As our spectrum policy slowly moves away from broadcast and towards an Internet-based platform, media will become more diverse and democratized due to the lower barriers to entry and ease of distribution. If the FCC moves away from the trustee and auction models towards unlicensed cognitive radio, the amount of wasted spectrum could be dramatically reduced, giving TVBDs the bandwidth to truly become mobile broadband devices. If cognitive radio can compete effectively with licensed wireless and wireline providers, it can exert pressure on the incumbents to provide better, cheaper and more diverse services.

  • But any wireless will do that: the issues is why we should use those particular frequencies in a condominium rather than look elsewhere for spectrum to use for WISPs.

Despite the fanfare, the particular kind of cognitive radio approved by the FCC last year has a number of internal and external constraints that will hamstring its impact on competition, consumer welfare, media diversity and spectrum efficiency. Preserving a preference for legacy technologies despite a more efficient and flexible replacement is a fundamental error that will slow both the development of cognitive radios and the proliferation of broadband generally. Broadcasters should not be able to crowd out entrants using the legitimacy of the license. The FCC must revisit its spectrum policies and reevaluate its public interest commitments.

The Potential of Cognitive Radio in the TV Bands

Cognitive radios challenge several central assumptions about spectrum allocation and media diversity. (Mitola and Maguire in 1999) Simply put, cognitive radios enable more speech in less space, reducing the need to artificially stop competition to coordinate use or encourage investment. Instead of being limited to a single channel, or requiring the channel to be altered by the user, cognitive radios use spectrum sensing software to determine what is the best mode of transmission. The IEEE 802.22 working group began to develop a Wireless Regional Area Network (WRAN) standard in 2004, the first such standard based on cognitive radio. (Cordeiro & Challapali).

[Mesh] Cognitive radios based on the 802.22 standard are capable of creating wireless mesh networks to improve penetration and network performance. This improves upon the hub and spoke model, enabling devices to communicate with each other directly as well exchange data with a central hub. As we have seen with mesh WiFi? networks based on the 802.11s protocol, such as the Meraki, mesh technology may allowing for more innovative and more resilient deployments, at lower cost.

Although the new ATSC digital signal enabled smaller channel width, and higher quality audio and video than its analog predecessor, DTV will has limited utility compared to more advanced, packet-switched technologies. The ATSC protocol is still premised on a hub-and-spoke, push distribution model in which signals are constantly broadcast, without regard to the presence of receivers or preferences of users. This gross over production of DTV signals relative to demand sharply reduces the total supply of bandwidth, especially in larger markets with more stations, potentially wasting billions of dollars of bandwidth that could be sold, or returned to the unlicensed commons. Broadcasters have a number of incentives to overproduce content in order to fill 24 hours of programming and maximize their ad revenue, while their duty to the public to produce educational and local programming is miniscule or, in many cases non-existant. (cite) Despite this moral hazard, there is currently no interest at the FCC to police this conduct in order to protect the other uses.

In the TV Bands today, interference is avoided primarily via power limitations and buffers of spectrum that prevent two signals from getting too close; The empty guard bands – or “white spaces,” as they are commonly known – were created out of necessity. But due to their flexibility, cognitive radios are able to squeeze into these non-contiguous chucks of spectrum, even during mobile use, when the available spectrum may shift from minute to minute. Furthermore, because TVBDs utilize packet-switching protocols, the information sent and received by a devices are at the control of the user instead of the hub. This reduces unnecessary traffic when a device is turned off, is out of range, or when it is receiving data from another source. By sending only what one asks for, there is more left for everyone else. Furthermore, because TVBDs utilize the flexible and impartial Internet Protocol, otherwise incompatible uses can be made work together, not only shrinking the size of buffers, but actually exchanging packets as nodes in the same mesh.

But it is the ability to send information, as well as receive, that is TVBDs biggest advantage over broadcast. If every receiver were also a producer, there would be a paradigm for media diversity, consumer welfare and public discourse. Communications policy has actually been moving towards greater consolidation in the last generation, culminating in the relaxation of ownership requirements earlier this decade. If traffic were allowed to travel upstream as easily as downstream, broadcasters would be forced to essentially become common carriers of packets, in a public square where anyone with an opinion would be allowed to speak.

The Battle in Washington over TVBDs

The idea of using cognitive radios in the empty TV bands has existed for sometime, but really began to gain traction in 2006, when a group lead by Google, Microsoft, Dell, HP, Phillips, Earthlink, and Samsung formed the White Spaces Coalition. The Office of Engineering and Technology (OET) was intrigued by the idea, and selected the emerging 802.22 standard as the basis for TV Band Transmissions. The Coalition argued that unlicensed networks in the TV Bands could be harnessed to raise a floor of wireless Internet connectivity that would cover much more of the nation than WiFi? ever could.

Supporters claimed that the performance of TVBDs would place competitive pressure on incumbents in wireline and licensed wireless to improve the overall quality of broadband offerings. Initial estimates were as high as 10MB/s, even at distances up to 30 km. (cite) If these claims held up, TVBDs would not only challenge 3G and 4G wireless services, but could become a legitimate ‘3rd pipe' to the home, solving the last mile problem due to its long range. These performance claims were largely theoretical, and based on several shaky assumptions but even with bandwidth sufficient to support only basic applications, such as email and location aware services, such a network would transform the way we interact with our environment and with others.

White space advocates were also able to sell the FCC on the unknown innovations that an unlicensed network in the TV Bands could provide. Taking a cue from earlier unlicensed technologies, the white space spectrum was pitched by Larry Page and others as 'WiFi on steroids.' Opening the 'junk' bands to unlicensed innovation led to the creation of a multi-billion dollar market of products we use every day. A TVBDs are actually said to have several advantages over WiFi? : Spectrum under 1GHz has much better propagation characteristics, allowing for longer range and high bitrates.

  • But with different interference characteristics and larger optimal antenna sizes. This would be a good place to discuss a little more fully the nature of the technological discussion between optimists and pessimists, or between those who favor using TV bands and those who prefer trading that spectrum for spectrum somewhere else.

Combined with the flexibility and efficiency of cognitive radio, it is proposed that unlicensed use of the TV Bands would fundamentally challenge assumption of what spectrum was capable of. The WiFi? example stands in starkly against the assumption that a private actor will only have enough incentive to invest in spectrum if they are given the right to exclude. Given the numerical scarcity, licenses often convey market power which often allows them to act anticompetitively. Consumers would benefit greatly if licenses were only issued or renewed when absolutely necessary

  • This is a powerful argument summarized too quickly to achieve its full impact with a reader just learning the subject.

On the other side of the table, the National Association of Broadcasters (NAB), wireless microphone manufacturer Shure, live event groups and the major sports leagues joined in opposition of TVBDs. F.U.D. was aggressively used to earn generous concessions to protect from legitimate concerns of coordinating uses, but even further to suppress even the suggestion that broadcast should have to accept some cost in order to maintain their licenses. These critics claimed that TVBDs would destroy our system of free broadcast and wireless microphones by creating profligate and unpredictable interference with existing uses. CEO of Maximum Service Television David Donovan liked TVBDs to “millions and millions of… interference causing devices, like "germs," [spreading] throughout America with the ability to attack the TV receivers in people' s homes, apartments, hotel rooms, hospital rooms, dormitories, etc. (Notice of Ex Parte Communication ET Docket Nos. 04-186, 02-380) Broadcasters’ sudden concern for the needs of the public was clearly pretextual for their own interest in keeping the costs of compliance as low as possible. But the meme of ‘harmful interference’ was pushed so aggressively that many more important and complicated issues were obscured, and TVBD backers were forced to make a number of concessions to quell opposition.

The OET continued to test prototype devices from several manufacturers to get to further test TVBDs ability to sense and avoid broadcast and other licensed wireless signals. In the second round of field testing by the OET, it was demonstrated that long-range cognitive radios were capable of identifying television signals via spectrum sensing. Although the tests did not attempt to gauge transmission speeds, the Commission stated in its October 15, 2008 report that “at this juncture, we believe the burden of ‘proof of concept’ has been met.” (Evaluation of the Performance of Prototype TV-Band White Space Devices Phase II) But although ultimately the incumbents lost the interference battle, they successfully framed the debate in a way that put the entrant on the defensive, and assumed the continued legitimacy of their licenses and their protocol with .

November 4th 2008

On Election Day, the FCC voted 5-0 to allow both fixed and portable TVBDs on the TV Bands on a secondary basis within approximately 200MHz of spectrum scattered between 150 and 862 MHz. The this vote allowed the FCC’s experiment with cognitive radio to move forward, but their potential of this experiment was artificially capped due to fears of interference stoke by broadcasters, and the reluctance of taking on license holders. Mobile TVBDs are allowed to broadcast at a maximum 100mW, and the devices also employ automatic power controls that will ensure no more power is used than is necessary. There was no attempt to quantify the remaining utility of broadcast television to consumers, nor was there an attempt to balance this utility against the potential of TVBDs. While the legacy technology continues to receive a huge subsidy in their renewed licenses, TVBDs seem to be a merely tolerated use. TVBDs are forced to compete with one arm behind their backs, while no burden of any kind was to be imposed on incumbents to accommodate TVBDs.

[Determining empty channels] The mechanisms for determining which channels can be used where the TVBD is located are conservative and burdensome. To begin, mobile TVBDs are equipped with spectrum sensing technology that considers 'occupied' channels that report signal strength higher than a given level. The Second Report and Order set this threshold at -114dBm, 1000 times weaker than the minimum required to get a DTV signal. (cite) A threshold of -107dBm was recommended by the IEEE as more than reasonable, but license holders successfully stoked fear over interference. (cite)

Although spectrum sensing was shown to operate correctly in the second round of testing, incumbents successfully pushed for the requirement of an additional safeguard to protect broadcast signals, pushing spectrum sensing to the back burner. The FCC has required that a geolocation database be created and maintained to catalog all licensed uses and their protected contours. . The cost of building and maintaining this database will fall on the entrant, not the incumbent. This belt and suspenders approach may significantly raise barriers to entry and prices, while reducing coverage area and battery life. Devices that are unable to access the database will not be allowed to transmit, even if using spectrum sensing they can locate free channels. This move cuts into the benefits of cognitive radio by forcing control of the network back towards the center, where powerful broadcasters may have a better shot at gaming the system. It is yet to be seen exactly how the database will operate in practice, but there may be reason to fear false entries or other abuses.

Although 200MHz is quite a lot of bandwidth, it must be remembered that TVBDs are sharing space. Although TVBDs are given “secondary priority,” broadcast, wireless microphones, cable television head ends, medical telemetry devices, translators and low power television are all given priority in the TV Bands. (Second Report and Order) Each is given a protected footprint similar to broadcast, multiplying the ‘no-transmit zone’ and further limiting the unlicensed spectrum for TVBDs to pull from. (cite Second R&O re database). The natural effect this ordering scheme are likely to produce grossly over-exclusive 'no-transmit' contour that could drastically reduce the available spectrum, which would keep TVBDs from reaching viability. As we are getting a first look at the data that will fill the database available channels database, many more channels are reported as occupied than was originally expected: It appears that original estimates did not include licensed uses other than broadcast. (see, e.g. In most major markets, only one or two channels will be available, far too little to provide anything close to broadband access.

Because TVBDs will be forced to work around multiple uses incompatible with the IP platform, many of the gains that cognitive radios could enable are squandered by incumbent uses that were grandfathered into this new scheme. This unequal sharing arrangement is a critical distinction from existing unlicensed bands such as WiFi? , but has been largely overlooked. Without the power to exclude wasteful, licensed users, the WiFi? on steroids analogy simply does not hold up. TVBDs will only be successful if more spectrum is returned to the commons for them to utilize.

  • You should have a conclusion here: the FCC gave with one hand and took back with the other.

Unresolved issues that may limit the impact of TVBDs

The Commission referred to the Second Report and Order as a 'conservative first step' to test out this new technology; There are a number of unresolved issues that will have a significant impact the development on TVBDs. The FCC will next revisit the Second Report and Order next year, but as of now, the path to market is far from clear.

Performance under real-world conditions is perhaps the most critical factor for TVBDs. What types of business models will arise to provide service, will largely depend on which other service they can offer a replacement to. Using one 6Mhz channel, the 802.22 standard is theoretically capable of a maximum bit rate of 19 mbit/s at 30km, or over 2 MB/s. (cite) But such theoretical assumptions should not be the basis for evaluating impact, and it does not seem that potential entrants are expecting such performance in real life. No significant field tests were attempted by the FCC to gauge performance in real world conditions where many different users would have to share bandwidth from available channels.

The 802.22 standard will continue to develop into a more powerful protocol, but without further details speculation on performance – and predictions of the competitive impact – is premature. Using channel bonding, MIMO or other techniques, the devices may eventually be capable of better performance in later generations. Needless to say, it is highly doubtful in the near future that a TVBD network will be able deliver mobile “broadband” service to a large number of people at once, especially given the new legislative definitions.

  • This graf sounds as though you didn't write it. You need to unpack more here for readers who don't follow this stuff habitually.

Because of the negative correlation between the number of broadcast licenses in a given market and the number of vacant channels, there will be significant variations in performance between urban and rural markets. In urban markets, where most broadcast channels are spoken for and many additional applications such as Commercial Mobile Radio Service may be present, there may be insufficient spectrum for devices to function at all. In this respect, TVBDs will fail to become a free or low cost way to access the Internet for disadvantaged groups in urban markets.

White Space networks will have much greater bandwidth available in rural markets. Unfortunately, rural deployments face a number of additional barriers to deployment. Due to their distance from an Internet backbone, and between their scattered customers, WISPs have found it very difficult to make deployments that make economic sense. TVBDs will lower the cost of deployment, but given the small customer base, it is unclear that there will be enough incentive for large scale rural deployments. In Both rural and urban environments, it is unlikely that TVBD will be able offer a full-featured service, either as a first option or as a replacement for existing broadband service.

Another hurdle to these devices entering the market is the device approval process. All devices must be approved directly by the full Commission, instead of being delegated to a committee through the normal process. (Second Report & Order) Giving each of the 5 Commissioners veto power could easily be abused for political reasons. Given the skepticism shown these devices during testing, there should be serious concerns that devices will meet the “proof of performance” standard. The threat of a veto could be used as leverage to extract additional pro-broadcast concessions, or to completely block any device based on the technology. The inability to rely on objective standards to predict device approval may severely undercut the willingness of device manufactures to invest in research and development.

Finally, it should be clear that unlicensed spectrum is certainly not synonymous with free Internet access, although the two seem to have been conflated in the public debate. We still know very little about what sorts of companies will provide TVBD service or how their business models will be structured. There will be considerable infrastructure costs to set up TVBD networks and keep them operating. Although this cost should be relatively low compared to other technologies, the cost of covering large areas, let alone a nation-wide network, will need to be recouped. One of the main problems is uncertainty. Because of the inability to exclude others from this spectrum, it will be difficult to make credible quality-of-service obligations to customers. Although many are used to nationwide cellular plans, it would be difficult to ensure that a mobile device will work at all.

The major players have yet to tip their hand, and no significant new players have emerged, but we soon should have an idea of which direction they are headed. Initially it was speculated that Google would provide a no-cost, nation-wide ad-supported network, but they have since stated they have no interest in providing wireless service. Free service may still be provided in order to subsidize some other revenue stream, but we should be wary of networks that exchange no-cost access, for unfree service.

In rural markets, where WISP companies are already established, there may be a smoother transition. But there is still the problem of reaching a backbone, and deploying enough base stations to cover wide areas. There will certainly be lower deployment costs, but it is unclear if they will be enough to cause a sea change in the viability of wireless deployments.

Another idea for operating TVBD networks proposed – and patented – by Google would enable live auctions to take place for local networks to offer service to TVBD devices on a short term basis. (Patent App. # 20080232574). This would allow a patchwork of networks to come together to approximate nationwide service, lowering the cost on any individual network operator, while simultaneously allowing customers to hop from network to network.

Local government actors may also join the act, and create city-wide TVBD networks that would serve both private and public functions. Municipal wireless networks have been attempted in the past without much success, but the longer range of TVBDs might make such a model viable on a large scale. Public-private partnerships have also been discussed, where the municipality would be the “anchor tenant.”

External Factors affecting the development of TVBDs

As stated above, the crowded TV Band is perhaps the biggest bottleneck to the development of TVBDs. Unfortunately, regulators missed a huge opportunity to move away from the licensed model and towards communal control of the network. Instead of reevaluating the continuing need for broadcast licences an era of media saturation, Congress simply renew existing licenses (at no cost) in the DTV transition. Although the switch to digital increased picture quality, preserved free broadcast and generated $19 Billion at auction, the primary beneficiaries were broadcasters, not consumers. It was also a huge transfer, perhaps as high as $70 Billion, from the public to private (Bob Dole dubbed "Telecom Giveaway,” Cong Rec. S135 (daily ed. Jan 10, 1995)) This reflects a bigger problem at the FCC of rubber stamping license renewals, to continue to do in the face of alternate media and multi-use platforms like the Internet is particularly troubling. (cite)

Despite this continuing bias, the end of broadcast is rapidly approaching. There are nearly 2,000 full power and 3,000 low power stations, and a significant number of them may not make the transition at all, moving to other media or ceasing operations all together. (cite). The amount of spectrum occupied by broadcasters is expected to decrease in the coming years, as ad revenues and viewership move to other media. It will be important to watch how long local broadcasters can hang on.

This trend perhaps explains why the White Space Coalition so quickly accepted that they would have to back off the more contentious aspects to make TVBDs less threatening to incumbents. Perhaps the WSC simply believed that as long as they were able to deploy cognitive radio in some form, they could afford to lose in the short term. This would seem to jive with Google’s position in the 700MHz auctions and their participation in developing the geolocation database.

Though TVBDs share spectrum with broadcasters, perhaps the more relevant way to assess TVBDs potential for disruption is to compare their performance with close substitutes such as LTE, WiMax? , cable and Fiber-To-The-Premises. As long as broadcasters hold onto their licenses they indirectly benefit these competitors wireless providers by starving TVBDs of bandwidth. In the coming years, these technologies will continue to evolve, which should bring about greater performance, more complete coverage areas, and more diverse service offerings. Licensed wireless technologies will significantly increase total supply of wireless bandwidth, which should tend to lower the cost of bandwidth. But the performance of these technologies will greatly outstrip even generous estimates of TVBD bitrates. Making TVBDs unlikely to be seen an unlikely substitute for tomorrow’s bandwidth needs. Differentiation may allow the continuation of the anti-competitive behavior that exists today such as asymmetric upstream and downstream bandwidth and throttling of peer-to-peer.

  • Surely this would be the place to point out that the network operators who want to replace broadcasting with their own content-delivery models are interested in this technology only if they could own and control it.

Given the many obstacles and unresolved issues facing this technology, it will be several years at best before TVBD networks have a significant impact. In the short term, TVBD service providers may only be able to peel off lower-bandwidth uses and users, but it will be too slow and unreliable for many key applications such as VoIP? or streaming video. This will blunt the impact of TVBDs on consumer welfare and media diversity. The incumbents may have to react to this competition to stave off substitution, but they'll only push themselves to innovate as much as they have to. Unless TVBDs can begin to offer dependable bandwidth in appreciable quantities, it will only service as a supplemental service, not a replacement.

Additional steps must be taken, both to develop cognitive radio and to ratchet up the pressure on broadcasters, either to force them to make the highest use of their licenses, or to force them to give their licenses up. The meme of harmful interference pushed by broadcasters is a false argument that assumes the superiority of Broadcast. All incompatible uses ‘interfere’ with one another; We must evaluate which uses create the most externalities relative to the value they offer, and embrace those that allow for the most robust performance, and diverse applications and users.

The FCC’s Public Interest Obligations

The core task of the FCC in evaluating its spectrum policy is to balance necessity of licensing against the harm of denying access to the rest of society. By allowing certain actors a voice and denying that voice from all others, speech is suppressed in derogation of the constitution. The FCC has long relied on the trustee model for spectrum allocation, which assumes the public is best served when certain entities are given the exclusive rights to operate on a given channel. This was intended to encourage the licensee to invest in infrastructure. Licensing was only justified out of necessity. The trustee model was necessary when analog signals could not coordinate without physical separation.

Spectrum is a public resource. This scarcity meant that conveying market power on a few actors in each market was necessary to coordinate uses. The lucky few were seen as fiduciaries of the public’s interest in spectrum, but in practice broadcasters have shirked this duty and the FCC has not enforced it. Free spectrum licenses are huge subsidy to broadcasters, who not only get access to the airwaves, but are able to leverage a free license to sell advertizing. Whether granted to the highest bidder or a trusted broadcaster on a public interest theory, licenses can no longer be seen as the default model of allocation.

As the barriers to entry have fallen, and protocols became more efficient, the justification for denying access no longer holds. Where the new technology is in some way incompatible with existing uses, the FCC must attempt to balance the merits of each use, to decide how to allocate spectrum between them. Scarcity is much less of a concern today as it once was, although the Supreme Court has been reluctant to revisit its holding in Red Lion.Over the past few decades, digital technology and cognitive radio have altered the traditional balance between free speech and allocational scarcity. This means that we no longer have to suffer as much anticompetitive and anti-democratic harm in order to allow production and distribution of content. Wherever possible, the FCC should to retire, or refrain from granting, spectrum monopolies that are unnecessary to encourage investment and beneficial use. The burden should shift to licensees to justify their continue need for dedicated spectrum. There may be an argument that without market power, there would be a market failure and the programming currently on broadcast would cease to be produced, but that argument has yet to have been made. And despite the billions of dollars spectrum has fetched at auction, the profit motive of the FCC must be subordinated to their duty to promote the public’s interests in spectrum, and we should no longer assume the highest bidder will put it the highest use. Unfortunately, the FCC appears to be reluctant to abandon the trustee model. There has been a move towards openness with the open access requirements attached to the 700mHz C Block, but these practices have yet to be imposed upon existing licensees via the renewal process. The commission needs to step back from its preference for broadcast and allow these technologies to truly compete on the merits.

While licensing may still be defensible for certain applications, the right of a private party to exclude to all others from a public resource should be presumptively anticompetitive, absent compelling justification. Aggressive scrutiny of a licensee’s public interest obligations could also lead to more vacant channels. The FCC should resist claims by incumbents that only by allowing them to keep their licenses can the public interest be served. Incumbents are digging in their heels, not to protect consumers, but to protect the mode of distribution and spectrum allocation that has given them a huge advantage for decades.

It was a mistake to believe that cognitive radios can flourish when they are only allowed to use the spectrum their licensed competitors leave empty. Developing a single comprehensive protocol for the TV bands would have great implications for spectrum efficiency, create great network effects, and promote diversity. Cognitive radio and the Internet Protocol are broad and flexible enough to handle these tasks. The FCC should fully embrace that cognitive radios are a key technological advancement and that the benefits of broadcast are too narrow to justify license renewal. If cognitive radio technology can spread across these discordant uses, and fewer licensees are left to crowd the airwaves, unlicensed networks can proliferate and true spectrum equality may become plausible.

The FCC must stop subsidizing broadcast via licensing, and should not wait for broadcast to die a natural death. It will become increasingly clear that TVBDs represent a higher use of this spectrum than broadcast. The public benefit of broadcast is obviously limited to the number of consumers that use it. Nearly all broadcast content is retransmitted via cable or some other service. By the time the Digital transition is complete, Broadcast penetration is expected to be below 20%, well behind, cable, satellite or broadcast. (cite) Digital broadcast is a simply a blunt instrument to distribute Television to the few people who do not have other access to other media. The cost to society of supporting broadcast via license subsidies is no longer justified by the benefits.

Furthermore, for those people, the benefits of broadband/IP are broader, both in the variety of uses and the number of users. There are areas and populations that do not currently have access to alternatives but these are precisely those that will benefit the most from a robust unlicensed. When you look at what the people need, they need Because the benefits of broadband are so broad and so flexible, the FCC should focus not on protecting the 20% of households that depends on broadcast, but the 25% who cannot afford, or don't have an option to purchase, broadband access. If the The FCC must seriously consider the narrow benefits and the broad externalities imposed by digital broadcast. By providing more services and a greater range of devices on the same slice of spectrum, the FCC could better fulfill its duty to put the spectrum to its highest use.

People need broadband, not broadcast. Even if licensing survives, more value could be created if the incumbents were forced to internalize the cost they impose on all other who would use this spectrum. The allocation of responsibility proposed by the Second Report and Order completely ignores efficiency concerns regarding allocation of responsibility for avoiding interference.. If broadcast and other exiting users of the TV were distributed over IP using cognitive radio technology, potentially hundreds of billions of spectrum could be returned to the commons. It may eventually eliminate the need for private spectrum licenses all together, when we can achieve true equality in spectrum, ending the anticompetitive harm caused by spectrum oligopolies. Undoubtedly there has been a great amount of money expended on the DTV transition, but it is not obvious that despite these sunk costs, switching to a cognitive radio-only solution in the TV Bands would not be a net gain.

The government has a clear commitment to the promotion of information and culture through its spectrum policy. The FCC’s focus going forward should center on how to increase broadband access. Because the benefits of cognitive radio are so broad, each user can decide what their highest use is, instead of a corporate trustee. The FCC should refocus on subsidizing the growth technologies in pursuing truly universal broadband in this country.

  • This section is important, from my point of view, because it explains clearly what the Free World wants from the FCC, and I think it does so pretty powerfully. But it seems to be a disruption of the flow of this essay's argument, which is less theoretical and more practical. I'd hate to lose it, but I think it needs to be better integrated if it is to stay.

Strategies to increase Unlicensed Spectrum under 1GHz

The decline of broadcast will grow the total amount of unlicensed spectrum, but a more aggressive approach should be taken to ensure investment in this technology is at an optimal level. The FCC and Congress must fight to recall the market power that they have given to media oligarchs via licensed spectrum. By redistributing the broadcast subsidy to the commons, consumer welfare, the public interest and the constitution will all benefit.

As mentioned above, the economics of broadcast are in a poor state, but that does not mean the FCC has no role in policing private licensees. Although the FCC has never scrutinized license obligations very seriously, there are several hooks that may lead to a license being revoked. While note necessarily taking a hostile attitude towards licensees, as the FCC takes a more aggressive approach towards enforcement, it will raise the costs of compliance and accelerate the demise of stations not meeting their obligations.

The FCC could force broadcasters to keep up with innovations in spectrum technology, which should lead to a smaller and smaller signal footprint. Alternately, the FCC could mandate symmetry between upstream and downstream, or put measures in place to prevent discrimination against applications, limiting the ability of licensees to leverage their licenses into other markets.

Once the 'proof of concept' was accepted by the FCC, they have an obligation to research how the technology can be utilized to further the FCC’s goal. The FCC should conduct more Notices of Inquiry to see how the 802.22 standard can be integrated into integrated into the other uses of the TV Bands such as wireless mics. The IEEE and device manufacturers will likely do much of the heavy lifting, but the FCC (and Congress) should focus on ensuring that our laws and regulations embrace this technology.

One of the most important thing the FCC can do in the short term is begin the process of converting the various users of the Television bands to a platform based on IP. If a single standard and protocol can be adopted for all users of the television bands, this valuable spectrum can be harnessed in ways never seen before. Users of different devices will benefit from others employing cognitive radio technology as they all become nodes of a single mesh. If these other uses were made to use the same technology as TVBDs, Kaldor-Hicks improvements in consumer welfare could be gained. This will not come without cost to those parties, so the FCC must decide how such expense will be allocated. But as cognitive radio and IP can become the standard use of these bands, the costs of implementation will continue to fall. Furthermore, the network would be more robust, reliable as the mesh becomes more and more dense.

  • This is the direction Reid Hunt would have gone, doubtless, if he could have also ended the interlacing of digital TV signals. But IPTV makes less sense when the video being streamed isn't compatible with other digital video. So you need to revisit an old controversy.

And the efficiency gains are not limited to private licenses. The FCC can also play a greater role in encourage adoption by local, state, and federal governments. Municipal networks are a potential, as are public-private partnerships, with the city as an anchor tenant. Municipalities have a number of communications needs, from traffic monitoring to parking meters, to surveillance and public safety. To the extent governments can integrate this technology, more bandwidth will be left over for use by the commons. Eventually the government may indentify spectrum controlled by the government that could be “rented out” for unlicensed use using a process similar to the Google patent.


Unlicensed cognitive radios offer unique environment for investment in innovation, without the need to grant market power to a handful of trustees. This breakthrough allows spectrum to be more directly accessed by the public, the true owners of this resources TVBDs are just one form of cognitive radio and we should see further advancement as more money is poured into its development. Beyond that, it all depends on how powerful they are allowed to be. The Commission has not gone far enough to promote cognitive radio. The problems we face today in broadcast and telecommunicaitons will persist despite TVBDs, unless more aggressive action is taken.


Webs Webs

r6 - 22 May 2009 - 15:31:06 - EbenMoglen
This site is powered by the TWiki collaboration platform.
All material on this collaboration platform is the property of the contributing authors.
All material marked as authored by Eben Moglen is available under the license terms CC-BY-SA version 4.
Syndicate this site RSSATOM