DISA Turns to Analytics to Solve Spectrum Puzzle

The Department of Defense has a genuine “do more with less” problem on its hands when it comes to electromagnetic spectrum.

The spectrum–the range of frequencies that carry electromagnetic radiation, including wireless voice, video, and data transmissions–is essential to what DoD (and much of the rest of the world) does. “Every single system on the battlefield is reliant on the electromagnetic spectrum,” Bob Schneider, technical director of the Defense Information Systems Agency’s Defense Spectrum Organization (DSO), said at a symposium last month.

But the spectrum’s capacity also is finite, and it’s getting increasingly crowded, with more wireless devices, systems, and Internet of Things (IoT) tools being added every day. For DoD, spectrum capacity is effectively getting smaller, both figuratively and literally. The Pentagon’s Electromagnetic Spectrum Strategy, among other measures, complies with a mandate to give up an additional 500 MHz worth of spectrum for commercial use by 2020.

The problem is how to solve the puzzle of working in an increasingly complex and contested environment. “We have a data challenge,” Gregory Wagner, chief of DSO’s Strategic Planning Office, said at the symposium. “We perceive the need for data analytics in this space.”

The spectrum doesn’t just support military operations, but is a battleground itself for cyber operations and electronic warfare tactics such as disrupting the enemy’s communications and the use of the spectrum while fending off an enemy’s own efforts at disruption. DoD has even considered declaring spectrum its own domain of warfare, along with land, air, sea, space, and cyberspace.

The Defense Information Systems Agency (DISA) has known for some time that making the most efficient use of spectrum would involve spectrum sharing both among the military services and allies, and among other government agencies and the public at large. In 2015, DoD signed onto a collaborative framework with the Department of Commerce aimed at more efficiently sharing communication channels.

A key to spectrum sharing is spectrum management, which DISA points out must account for the three “Big C’s.” The first C stands for “congested” because of all the devices using it (many cellphones, for instance, operate in the 1755-1780 MHz band where DOD relocates its systems). The second C stands for spectrum “contested” actively by adversaries. And finally, the third C stands for “constrained,” because of the regulations about how DoD can use spectrum at home and abroad. The department must be able to manage its use of the spectrum within those confines, and be able to adapt to changing needs on an on-demand basis, which is where DISA’s call for data analytics comes in.

Wagner noted that demands on spectrum aren’t static–what’s needed today might not be what’s needed tomorrow. DoD also needs to be able to react to interference when systems are sharing spectrum frequencies. Analytics–in the form of algorithms, automation and data visualization–can support techniques that address how spectrum can be shared and for how long on a changing basis. “The demand signal says we are going to have to do more and more of this [spectrum sharing],” Wagner said. “We need to drive the time scales from the manual to the automated [and] from the calendar to the machine.”

Analytics, for example, can be applied to assessing interference, determining whether it is intentional or unintentional, or resulting from atmospheric or weather conditions. “One of the biggest problems with spectrum data is there is a lot of ambiguity in the structure and the values,” said Elizabeth Park, a DSO electronics engineer. “We are looking toward machine learning to help make sense of our data and correct it.”

DISA also is looking to improve visualization of spectrum data, for instance applying 3D gaming technology to get a clearer picture of the complex spectrum environment to support planned decision-making.

Recent