Rationalizing Government’s Spectrum Use

This Thursday, June 27, the House Subcommittee on Communications and Technology will hold a hearing on government spectrum use (titled “Equipping Carriers and Agencies in the Wireless Era,”) at 10:30 a.m. Invited witnesses include two from the government side, Teri Takai from DoD and Karl Nebbia from NTIA, and two from the civilian side, Dean Brenner from Qualcomm and  Chris Guttman-McCabe from CTIA.

According to the hearing memo, the questions to be asked will include:

How can Congress meet the needs of Federal agencies while addressing carriers’ spiraling demand for spectrum in the age of the smartphone? Americans increasingly expect—and our economy increasingly depends on—the ability to access information, enjoy content, and conduct commerce from a mobile device nearly anywhere in the country. And with most of the “low-hanging fruit” exhausted, the conversation is once again turning to spectrum currently used by Federal agencies to fill the breach. But government reliance on wireless technology shows no sign of diminishing, either. What tools are available to maintain and even improve Federal agencies capabilities while freeing spectrum for commercial use? How much would various approaches costs, and how long would they take to implement? What progress is being made today? What steps might Congress, agencies, and the private sector take to facilitate the process?

We can bet that there is going to be some lively discussion over the question of just how much spectrum the government currently controls, where the answer is 60% of the most valuable spectrum in the most crowded places, and 20% in rural areas. This hearing wouldn’t be necessary if the PCAST Committee had done a more credible job of analyzing government spectrum use last year, but in their defense it was almost impossible for them to dive into the details of government spectrum use because much of it is classified and PCAST members don’t have security clearance. So PCAST did pretty much the only thing it could do when denied essential details: produced a big-picture plan with almost no specifics proposing a way of sharing access to spectrum without knowing how it’s presently used. PCAST simply proposed the use of an authorization database around government spectrum to permit occasional use by civilians.

This was a big disappointment to the cellular industry and to those of us who use mobile devices as it prevented us from reaping the kind of spectrum bonanza the nation as a whole gained when television was converted from analog to digital and some 500 Mhz of spectrum was transferred from TV broadcast to mobile and unlicensed broadband. This first Digital Dividend was substantial, and a second one would have been even larger and more beneficial since the government controls more spectrum for its own use than TV did before the DTV transition.

DoD is still dragging their feet on spectrum efficiency and reassignment. The limited evidence we do have suggests that there has been a pattern of purchasing in military systems in which basic engineering principles such as design modularity and spectrum agility have not been honored. At one meeting I attended on spectrum policy last year, a PCAST contributor asserted that replacing the radios in some military aircraft with others using more up-to-date technology would require a complete tear-down and reassembly of the planes, at a cost of several million dollars.

DoD is still reluctant to share details, but it’s reasonably clear that government systems can benefit from upgrading and repacking every few years in much the same way that civilian ones do. If government systems observed modular design, this can be done on a routine basis, sometimes by a software upgrade.

DoD would naturally like to retain control of as much spectrum as they can, because they have enough vision and imagination to see possible uses on the horizon that will enable them to do their jobs better, just as civilian visionaries do. So the sharing notion really is essential, even if it’s utterly unclear how to make it work in an equitable way between two partners when one insists it’s more equal than the other.

In a perfect world, each band of spectrum in each area would have one manager, and that manager could ensure that users get the access they need by interleaving packets, as the Internet does and as cellular networks do. If PCAST had taken a more sober view of “sharing” they would have recommended this approach rather than the essentially dysfunctional “dynamic/cognitive” sharing approach they touted. In the wireless world, dynamic/cognitive sharing is more often than not a small network technology that doesn’t scale to large networks. But PCAST relied more on advocates of dynamic/cognitive from the academic world, many of them with no technical background, than on experienced network engineers and operators from the business world.

There are three specific issues the subcommittee can and should address:

1. Immediately reallocating the 1755-1780 MHz band for commercial use. This brings the US into alignment with the rest of the developed world. Per ITU’s band plans, 1755-1780 MHz is a commercial band that pairs with another slice for upload/download. Six of the 8 G-8 countries and 17 of the G-20 countries use it this way, it’s not special to DoD, and the carriers are willing to pay the government’s relocation costs.
2. On that point, auctioning government spectrum can provide the funds to relocate government users into alternate bands and buy them newer, more efficient equipment in general. Using auction revenue to pay for these upgrades benefits everyone and doesn’t strain the sequestered budget.
3. Moving government systems (and unlicensed systems) into bands above 3 GHz resolves the tension between mobile commercial systems and others.  The spectrum between 400 Mhz and 3 Ghz is uniquely suited for mobility, and many of the current and anticipated uses are either stationary (like backhaul) or nomadic (like Wi-Fi.)

Given that the mobile platform is now a key input into a wide array of applications and systems, it’s appropriate to recognize that true “beachfront” spectrum for mobility ranges from 400 Mhz and 3 Ghz and should be allocated accordingly. There’s a ton of spectrum above 3 GHz for sharing experiments, unlicensed, military and other purposes, so there’s not much of a rationale for the current allocation.

Ultimately, we need to have a spectrum custodian who can make reallocation decisions without undue pressure from the military, but that’s a steep hill to climb. For the moment, the ball is in Congress’ court.