DARPA's Software Defined Radio (SDR) Hackfest Creates Solutions For Spectrum Challenges

SDR community answers call to explore interplay of radio and drone technology during weeklong event

The DARPA Bay Area Software Defined Radio (SDR) Hackfest came to a close on Friday, November 17 at the NASA Ames Conference Center in Moffett Field, CA. During the weeklong event, over 150 members of the SDR community came together to discuss, innovate, and ideate around the future of software radio technology and its potential to address challenging communications issues that are emerging due to the increasingly congested electromagnetic (EM) spectrum and the proliferation of wireless-enabled devices.

“The DARPA SDR Hackfest was created to engage a growing community of SDR developers and enthusiasts from a diverse range of backgrounds and I believe we accomplished that mission,” said Tom Rondeau, a program manager in DARPA’s Microsystems Technology Office (MTO), who led the event. “The Hackfest provided community members with a place to interact with experts and explore new ideas around the potential for the technology, and our speaker series at the event challenged attendees to contemplate everything from the trajectory of the UAV industry to the challenges we must address to ensure the free and open sharing of software.”

Throughout the Hackfest, eight pre-qualified teams from academia, industry, and the hacker- and makerspace-communities worked together to develop solutions to specific “Hackfest Missions.” Presented to the teams on the first day of the event, the three Missions amounted to problem sets designed to examine how SDR could be used to solve communications, computing, and control challenges at the still-uncharted intersections of cyber and physical technologies. Each Mission focused on the communications link between ground stations and unmanned aerial vehicles (UAVs) or drones. These links are susceptible to breaking due to interference, EM congestion, or other communications issues. In developing the Missions, DARPA’s Hackfest organizers consulted with groups of military operators to ensure that various real-world scenarios involving radio and UAV technologies would be included. As such, it is possible that some of the solutions to these communications challenges that emerged during the event could progress toward real-work applications.

“The future use of drones–whether on land, in water, or in the air–will require far more coordination of the communications channels among all involved users. Trading off resource constraints of spectrum, processing power, and security risks, the Hackfest challenged teams to think through the whole stack from the application layer down to the physical,” said Rondeau.

Hackfest teams were asked to tackle one of the three Missions throughout the week. The goal of Mission One was to find a way to enable communications between a ground station and a UAV with a blocked or obstructed communication pathway. Teams were tasked with using a second UAV to create a relay link for the ground station to the obstructed UAV. This first Mission simulated real-world circumstances in which communication links can be denied for any number of reasons–from physical barriers to deliberate or accidental EM interference. Two of the participating teams took on Mission One, leveraging a number of open-source technologies to extend communications between the ground station and the obstructed UAV.

“YeS DR, a team of cybersecurity experts from Parsons, added an authentication method to their solution that helped certify who communication packets were intended for–a unique approach that played to their strengths as cyber experts,” said Rondeau. “Team DROGON also attempted Mission One. Bringing experts from Raytheon BBN Technologies and SSCI to the Hackfest, the team took full advantage of the provided code by using the Linux kernel’s BATMAN (Better Approach to Mobile Adhoc Networking)–a routing protocol that intelligently distributes information across a network—to rethink the mission as a network of nodes instead of separate ground stations and UAVs that were given different tasks.”

In Mission Two, teams were asked to show that control of a moving UAV could be repeatedly transferred between multiple ground stations. This Mission was designed to mirror scenarios where detailed flight work is required but a pilot’s visibility is degraded or limited due to, say, environmental conditions or changes in geography. This type of scenario is seen in civilian or disaster response operations where drones are employed. To jointly address Mission Two and One, one team sought to create a solution that considered every node in their system, whether a ground station or UAV, to be a network.

“Team Platypus Aerospace from Aerospace Corporate sought to address scaling problems beyond two or three drones by building a full mesh network solution with authentication and built-in encryption,” said Rondeau. “The solution the team developed to address both Mission One and Two could have applications for large swarms of drones where authentication of units within a swarm, as well as the messages passed between them, needs to originate from inside the swarm itself or its command and control station.”

Mission Three challenged teams to find new applications for SDR-UAV systems through the integration of sensors and the seamless transfer of information between a UAV and a ground station. One goal of the mission was to create a solution that could help improve how ground stations and UAVs interact, whether by reducing signal latency, allowing for the management of multiple drones without confusion, or some other means.

Throughout the Hackfest, four teams tackled Mission Three from a variety of angles and developed a range of solutions using onboard sensors. “Texas Radio Terminator, a team of graduate and undergraduate students from Southern Methodist University in Taos, took an approach of mixing sensors and feedback from the device in ways that should lead to better decisions about how to communicate and coordinate tactics with drone technology in the future,” said Rondeau. “Another university-led team, DeepEdge, used computer vision to recognize and track a face with the UAV. The approach used by this team, comprised of students and professors from University of California, Irvine and University of Southern California, demonstrated the possibilities of using autonomous tracking by making use of the application layer (computer vision) and the physical layer (SDR) for managing the computing and communications resources available. The team from Assured Information Security also took on Mission Three, creating new physical channels for each sensor, which provides a means to better manage spectrum resources.”

Bringing a diversity of backgrounds and expertise to the Hackfest, each team took on the Missions with their own styles of creativity. However, after a few days of problem solving, all teams came to the same conclusion–the true challenge didn’t reside within the individual technologies but rather in the task of bringing everything together. “Teams tapped on-site experts and resources, other teams, and the hacker-space community, generating the collaboration we hoped would be catalyzed by the event,” said Rondeau. “What we learned from this experience is that there is still more to be done to progress SDR, but the active involvement of a growing community will help energize efforts to keep things moving forward.”

With an eye on eking more value from the technology development that began during the DARPA SDR Hackfest, several teams have posted their innovations to open-source platforms and communities. Additional details about the DARPA SDR Hackfest can be found at https://darpahackfest.com/.