ITIF's Bennett says PCAST Spectrum Report Impractical
WASHINGTON - Richard Bennett, Senior Research Fellow at the Information Technology and Innovation Foundation, made the following statement on the report of the President's Council of Advisors on Science and Technology (PCAST) on spectrum policy, "Realizing the Full Potential of Government-Held Spectrum to Spur Economic Growth:"
"The report correctly concludes that the spectrum crunch is real and immediate, but fails to offer an actionable system capable of alleviating the spectrum shortage that has already begun to affect the 300 million Americans who rely on mobile handsets and smartphones today.
PCAST has offered an intriguing set of solutions to the problem of increasing the supply of civilian spectrum without disrupting military systems, but in doing so it has answered the wrong question. The urgent issue for spectrum policy makers is how to manage the ever-growing Federal appetite for spectrum without slowing economic growth and impairing the wireless services that consumers have embraced. While the PCAST report offers suggestions for better managing Federal spectrum, the question will need to be addressed in full elsewhere."
While the report conveys interesting information about research initiatives and intriguing data on current government systems, it fails to distinguish practical systems from speculative and unproven ones.
The system that it proposes would unduly burden American consumers and network providers by making their entire joint investment in wireless handsets and infrastructure obsolete. At the same time, it would protect all Federal users (primarily the military,) from any disruption to present operations, regardless of how inefficient current systems may be. This is not the proper balance.
The most astonishing claim made by the report is that "the traditional practice of clearing government-held spectrum of Federal users and auctioning it for commercial use is not sustainable." On the basis of this assertion, the PCAST report embarks on a thought experiment toward a new method of allocating spectrum which it terms "a new spectrum architecture and a corresponding shift in the architecture of future radio systems that use it [that] can multiply the effective capacity of spectrum by a factor of 1,000."
Claims of this magnitude should be supported by reams of empirical and analytical data, but the 162 page report offers no data at all to support its presumption that the auction system (which has been employed by the FCC only since 1994) is not "sustainable" or even to define the parameters of "sustainability."
The auction system doesn't need to meet the needs of spectrum users indefinitely, it only needs to provide a rational way to re-allocate spectrum from low-demand and low-value uses to those that the public values more highly until we have practical means of simultaneously sharing spectrum at the same times, places, and frequencies without undue interference. Our forthcoming report, "Powering the Mobile Revolution: Principles of Spectrum Allocation" describes these technologies and the timeline for their deployment.
The only system that has ever been effective at increasing the supply of usable spectrum is one that upgrades legacy systems, such as the old analog TV broadcast system, to up-to-date digital systems with greater bits/hertz information efficiency. The best of these modern digital systems conform to international standards such as LTE and Wi-Fi that foster the creation of supporting industries in silicon chips, handsets, base stations, antennas, and software.
The PCAST report rejects this approach in favor of new technologies that would favor interference tolerance over efficiency: "Reductions in the transmitted bits/Hertz reduce the interference footprint as a ratio of the communications range. Transmit waveforms should transition from maximizing the bits/Hertz in scarce spectrum to instead optimizing for spectrum reuse." Spectrum experts will naturally take issue with this finding, which is also not supported by evidence.
PCAST places enormous faith in the ability of geo-location databases to improve the usability of spectrum, consistent with the proposed White Spaces system that will rely on such databases when deployed. The White Spaces system enables fallow spectrum to be harvested and put to productive use, just as Wi-Fi enables consumers to operate their own wireless networks at home and in the enterprise. But these systems are a complement to commercial wireless networks rather than a replacement. In areas where there is no fallow spectrum in the frequency and power ranges that can be used by consumer devices conforming to international standards, no database or opportunistic access system can supply it.
It's important for legislators and regulators to ensure that consumers have an adequate supply of spectrum for short-distance Wi-Fi networks which are also under stress in many areas. The White Spaces system is likewise a worthwhile system that must be allocated sufficient spectrum to either succeed or fail in real operational settings. But neither the proven value of Wi-Fi nor the potential value of geo-location systems warrants PCAST's rash desire to put all of the nation's spectrum assets in one basket. Advances in technology are often messy and disruptive, so it's much more sensible to continue pursuing a multi-faceted strategy that allows technologies to compete on the basis of the value they offer consumers than to tilt the scales in favor of one and only one system.
Geo-location databases are a very important tool for managing unlicensed spectrum; they offer a means of forcing obsolete devices such as the early Wi-Fi adapters off the air in favor of better and more recent systems. They permit the rapid deployment of networks in emergencies and in unserved and underserved areas. Very importantly, these databases can be used to implement sharing policies that and can handle overload from licensed networks of various kinds. It's a mistake to assume, however, that these devices mandate any particular policies about spectrum allocation. Geo-location databases are t