Testimony Before House Subcommittee on Avoiding the Spectrum Crunch

Computing is undergoing a dramatic shift from fixed-location desktop and nomadic laptop systems to mobile devices, networks, and applications. In 2011, the number of smartphones sold worldwide exceeded the number of personal computers sold for the first time. Only half of Americans have smartphones so far, so the trend will continue for some time. One day appliances and other devices will come to have smartphone capability built in, so the number of “smartphones” will exceed the population many times over.

New users will use mobile social networks, among other applications. Last week, Facebook acquired Instagram, a photo sharing service with only 13 employees, for a billion dollars because Instagram had acquired 40 million users in only 16 months of operation. “Mobile Augmented Reality” is a new application category that extracts information from massive databases in the Cloud relevant to a user’s location, activity, and preferences; it moves video streams between the user and the Cloud. All of these applications require spectrum – the more the better – and as they’re truly mobile there are limited opportunities to offload their spectrum needs to short distance Wi-Fi networks. Spectrum assignments by regulators around the world have produced a highly fragmented system of relatively small assignments for a relatively large number of applications, as we see in the NTIA’s spectrum allocation chart.

We need to realign spectrum into a smaller number of larger allocations for general-purpose commercial networks because such networks have the proven ability to manage the demands of competing users and applications. In order to do this – a process akin to putting Humpty-Dumpty back together – we need to shift most government applications and all low-value commercial applications onto general-purpose commercial networks. This is where the 500 MHz recommended by the National Broadband Plan will come from, and the only way to get to a more realistic allocation of commercial spectrum. All spectrum assignments ultimately come from a common pool. Many government applications are critical for first-responders during periods of crisis. We have technologies that permit certain applications to get high-priority treatment on commercial networks. But commercial users also desire more spectrum during such events, so we have a policy conflict. This conflict was resolved by Congress through the creation of FirstNet, the public safety network operated by NTIA, but this is not a satisfactory solution. Ultimately, FirstNet operations should be commercialized, as soon as devices have been developed that allow trusted priority access policies. When we have such devices, the balance between public and government use can be specified by contract rather than by spectrum fragmentation.

Striking a balance between commercial and government use will remain a difficult policy problem until mobile network technology advances to the next stage. Ultimately, technology will enable reliable networks to support multiple simultaneous transmissions (many speakers at once) in the same spectrum, at the same time, and in the same location. Commercial network operators are motivated to solve this problem, but with the decline of America’s R&D giants – such as Bell Labs– funding for basic research is highly dependent on government’s contributions. Taxpayer money is better spent on such research problems than on building duplicate network facilities such as FirstNet. Advanced sharing will have tremendous military benefits as well, since it does not depend on cooperative regulators abroad.