WASHINGTON--The federal government is slinking away from a promise by President Obama to free up badly-needed radio spectrum for mobile users and the already over-taxed networks that serve them.
Just months after the publication of the National Broadband Plan in early 2010, the president issued a memorandum ordering the FCC and the Commerce Department's National Telecommunications and Information Administration to "make available a total of 500 MHz of Federal and nonfederal spectrum over the next 10 years" for mobile users.
The goal was to clear unused or underutilized spectrum the FCC could then auction off for use by an exploding mobile ecosystem.
Yet two years later, despite frequent assurances that progress is being made to head off a "spectrum crisis," almost no new frequencies have been offered for commercial use by either agency.
Now a report (PDF) published last week by the President's Council of Advisors on Science and Technology urges the president to give up trying. PCAST, which describes itself as "an advisory group of the nation's leading scientists and engineers" appointed by the president, recommends the White House issue a new memorandum that would dramatically reverse course, offering no new spectrum, now or ever, from government stockpiles.
Instead, PCAST is calling on the president to begin a long and technically-uncertain transition to a world in which usable radio frequencies would be shared by multiple users simultaneously. "The essential element of this new federal spectrum architecture is that the norm for spectrum use should be sharing, not exclusivity," the report concludes.
Adoption of this radical new approach to spectrum management would utterly change a system that has been in place since the sinking of the Titanic, and which has in recent decades generated billions of dollars through license auctions. The new model would, if adopted, have a massive impact on technological development, requiring, according to the report, "a corresponding shift in the architecture of future radio systems that use it."
As PCAST admits, implementation of large-scale shared spectrum would take years, add multiple layers of new federal bureaucracy, dramatically expand the role of the White House in on-going oversight of the airwaves, and require extensive new FCC regulations and enforcement for receiving devices such as smartphones.
At best, the PCAST plan is both technically and politically infeasible. At worst, it represents the latest step in a carefully choreographed campaign by federal authorities to shift attention away from their continued stonewalling on freeing up unused or underutilized spectrum.
Over the last two years, spectrum "sharing" has become code among federal authorities to stall for more time. The feds are playing a shell game, appearing to offer spectrum as ordered by the president but in fact offering nothing. And mobile users are already running out of both time and capacity.
A spectrum crisis looms
PCAST imagines a world in which government users would have primary access to spectrum currently assigned to federal applications, but where consumers could use those frequencies at times and in places where the government isn't actively communicating. While PCAST's vision of a shared "spectrum access system" is admirable, the kind of real-time spectrum reallocation it calls for is simply not cost-effective or in many cases not even possible given current technology.
The report itself acknowledges that PCAST's approach "represents a major evolution of existing spectrum management practices," and that "implementing it will not be easy and may take a long time." Its most optimistic prediction, which seems hopelessly naive of the actual pace of government change even on minor policies, is that putting the plan into operation would take three to ten years.
That's in part because there are no technologies in large-scale use, for example, that would allow mobile devices such as smartphones to dynamically change frequencies when users with higher priority need access, especially when the secondary user is in constant motion. According to Mary Brown, Cisco's director of government affairs, "no one in the world has ever tried to have a mobile technology share with some kind of federal system."
Meanwhile, "mobile networks in major cities are running close to capacity during peak periods already." That's according to Richard Bennett, a senior fellow with Information Technology and Innovation Foundation, a D.C.-based think tank. Barely a blip a few years ago, mobile broadband is growing at an astronomical pace. According to a report earlier this year (PDF) from the White House Council of Economic Advisers, mobile data traffic will increase another twenty-fold between 2010 and 2015.
Some of that capacity may come from more efficient use of existing network infrastructure and by increasing tower deployments and making use of smaller cell sites. "But much of it depends on more spectrum," Bennett said.
The FCC agrees, emphasizing its longstanding commitment to clearing and reallocating underutilized frequencies and auctioning them to those who would put them to higher and better uses. In response to the PCAST report, FCC Chairman Julius Genachowski told the group (PDF) that while spectrum sharing was a promising tool to improve the system, "Historically, our basic strategy has been to clear spectrum and reallocate it. This is a strategy that has delivered tremendous benefits for America."
Recently-confirmed Commissioners Jessica Rosenworcel and Ajit Pai have both called for more investment in sharing technologies, and quickly. According to Rosenworcel (PDF), "We must put American know-how to work and create incentives to invest in technologies--geographic, temporal, and cognitive--that multiply the capacity of our airwaves."
Regardless of the solution, time is of the essence. The 2010 NBP now seems conservative in its estimate that the insatiable demand of consumers for mobile broadband applications would quickly overtax existing network capacity.
According to the plan, 300MHz of added spectrum would be required by 2015, and 500 MHz by 2020. But different bands of spectrum have different properties, and not all bands are useful for mobile applications. Frequencies that are usable for mobile services are nearly all licensed for other uses. That means the only way to get more spectrum is to reallocate it from existing licensees.
The NBP identified two particularly promising candidates for reallocation. One is over-the-air television broadcast, in a steep decline for decades. Today, less than 20 percent of U.S. homes rely on broadcast, receiving programming instead from cable, satellite, and traditional telephone service providers. Increasingly, video is migrating to "over the top" Internet services, including Netflix, YouTube, and Hulu.
Earlier this year, Congress authorized the FCC to work with broadcasters who don't need some or even all of their existing allocations to conduct a new round of auctions that would license that spectrum to mobile broadband providers, generating an estimated $30 billion in auction revenue(PDF) to be shared by the broadcasters and the federal government.
Assuming the broadcasters cooperate, these auctions could go far toward closing the spectrum deficit identified by the NBP. But given the complexity of corralling the TV stations, these auctions won't result in new capacity for mobile users for several years at least -- long after the FCC predicts mobile networks will hit the wall.
The federal promise revoked
That's why so much hope is riding on the second potential source of reallocation -- the federal government itself. Today, federal agencies are the largest holders of spectrum, much of it in bands that are extremely useful for mobile services. And a great deal of that spectrum lies fallow, or is used only occasionally, often for systems that no longer require dedicated government use.
That was the view of the White House, in any case, in 2010, when President Obama ordered federal agencies to identify as much as 500 Mhz of federally-assigned frequencies that could be auctioned to mobile providers. The 2010 Memorandum ordered the NTIA to coordinate the federal efforts to release spectrum. (While the FCC manages spectrum licensed to commercial users, NTIA oversees spectrum assigned to the federal government.)
In March of this year, NTIA finally published its first detailed proposal for clearing unused or underutilized spectrum held by federal agencies.
The NTIA report began optimistically, noting that twenty federal agencies had identified almost 100 Mhz of contiguous, usable frequencies that could be cleared over the next 10 years. Federal users who still need the spectrum they have would be relocated to other frequencies that have less value to commercial applications.
But reading between the lines, the report noted agency estimates of relocation costs exceeded $18 billion, more than the amount likely to be raised by FCC auctions. Under current law, the FCC cannot hold auctions for spectrum currently licensed to government users unless it can raise more money than the costs of relocation. So the frequencies won't be cleared after all.
The lack of any details or analysis by NTIA, and the fact that the estimates were developed solely by the agencies themselves, led many here in Washington to suspect that bureaucrats were sandbagging, developing estimates in a way that ensured they would not have to relocate.
In no cases did the agencies identify systems that were or were likely to become obsolete in the near or long term, or which could be modified to require less spectrum. For example, multiple agencies operate separate video surveillance networks, many of them using outdated analog technologies. The Departments of Defense and Justice estimated that relocating their video surveillance, which by itself consumes 130 Mhz of spectrum, would cost over $5 billion.
Nothing in the NTIA report suggested that the less-sensitive of these systems could be migrated to commercial providers, or updated to digital technology. The latter solution, according to ITIF's Bennett, could free up as much as 75 percent of the currently-assigned spectrum.
Against this obvious stalling, the NTIA offered the consolation prize of sharing the spectrum between federal and commercial users during the transition, or in some cases on a permanent basis.
The proposed sharing arrangements, however, evaporated on closer reading. The agencies demanded, for example, that any commercial use be made secondary to unidentified federal requirements, with "clear regulatory mechanisms" to enforce their priority. According to the report, sharing would also require "acknowledgement by industry of its status with respect to potential interference from federal operations." Federal users, in other words, would be free to define the limits of acceptable interference, including in ways that make cost-effective sharing impossible.
Realistic limits of sharing today
It is true that some sharing between government and commercial users may be possible in the short term. A number of defense systems, for example, use their frequencies to communicate with satellite systems on a regular schedule -- perhaps only once a day or even once a month. So long as commercial networks stayed off those frequencies at specific times, these otherwise unused frequencies could be put into commercial use without relocating the federal systems.
In other cases, federal use is limited to certain locations, and commercial networks operating outside those areas could make full use of the frequencies without interfering.
This kind of geolocation sharing, as it is known, is already in use in many parts of the spectrum, and the FCC is now testing expanded forms of sharing in the white spaces between television channels.
The key to success for geolocation sharing is the creation and accurate maintenance of databases that specify where and when the primary user is operating. While already a significant obstacle for commercial sharing, as in the white spaces, many federal systems are classified. It is not clear if federal users will be able to provide the necessary information to make the databases usable.
Reading between the lines of the NTIA report, PCAST correctly concluded that it was unlikely the federal government would actually cough up any significant amount of spectrum for mobile users in the timetable ordered by the White House. But instead of calling out the agencies for obvious dawdling, PCAST doubled down, proposing an even more radical reliance on sharing technologies that are either untested or, in many cases, non-existent.
The more promising form of sharing emphasized in the PCAST report rely not just on databases but on intelligence built into transmitting and receiving devices. These include the use of "cognitive radios" that can sense when activity is already present on a given frequency, and "dynamic spectrum access" which can rapidly switch between frequencies as interference reaches beyond pre-determined thresholds.
The report hopes such technologies "will likely mature and become viable during the implementation of these recommendations." If so, their use could "multiply the effective capacity of spectrum by a factor of 1,000," which the authors argue is a "conservative estimate." But the report, here and elsewhere, offers no concrete data to support its wild claims. Nor could it.
While PCAST is to be commended for its vision of future technologies and their ability to make dramatic improvements in the efficiency of spectrum allocation, use, and minimization of interference, the kinds of technologies PCAST foresees simply don't exist yet. We don't know when they will come, and how well they will work. There is far too much uncertainty in technologies that "will likely mature" in the next decade to justify a complete rewrite of spectrum policy.
A call for presidential leadership
PCAST admits that their proposal "represents a major evolution of existing spectrum management practice, and that "implementing it will not be easy and may take a long time." In the best case scenario, all the authors can hope for is that "the long term direction outlined in this report can start to be operational in one to three years," with actual implementation of the new sharing model taking three to ten years.
Even that estimate requires wishing away the practical realities of reorienting the federal government to scrap a spectrum management system that has been in place for 100 years. Given Washington's profound inertia, even minor changes in policy can take years to implement.
Even if it were technically feasible, the certain delays in implementing this policy would stretch the PCAST time frame out by years if not decades. That's time we don't have. Mobile users will take little comfort from the grand ambitions of the president's advisors when their mobile devices can't perform even basic functions because the limited spectrum available to them has become saturated.
President Obama was quick to respond to the National Broadband Plan's conservative predictions of a mobile doomsday. He ordered federal agencies to start immediately the process of releasing spectrum that could be put to far more valuable use by mobile consumers. But two years later, nothing has happened. Government bureaucrats and their supporters have brilliantly thwarted any real prospect for federal spectrum reform at every turn.
Meanwhile, PCAST and the NTIA are fiddling while mobile Rome burns. And the White House is simply watching from the sidelines.
For the sake of our vibrant mobile ecosystem -- one of the few bright spots in our languishing economy -- we need real leadership from the White House. We need a practical plan for retiring outdated federal systems, and honest estimates of the costs of relocating those that are still necessary. We need sharing arrangements that can work with today's technology. And we need it now.
What we don't need are more two hundred-page reports touting unrealistic solutions, new regulations, and still more layers of bureaucracy. The president's 2010 memorandum promised to "unleash the wireless broadband revolution" by freeing up urgently needed spectrum for mobile users. But instead of innovation in spectrum management, all we've seen so far is the traditional genius of the federal civil service for resisting change. Mr. President, it's well past time to kick some butt.