GatorWings Wins $2 Million DARPA Spectrum Collaboration Challenge

GatorWings, has won the DARPA Spectrum Collaboration Challenge (SC2) – a three-year competition to unlock the true potential of the radio frequency (RF) spectrum with artificial intelligence (AI). GatorWings is a team of undergraduate students, Ph.D. candidates, and professors from the University of Florida.

DARPA hosted the championship event at Mobile World Congress 2019 (MWC19) Los Angeles in front of a live audience. SC2’s final 10 competitors and their AI-enabled radios went head-to-head during six rounds of competitive play. GatorWings emerged victorious, taking home first place and the $2 million grand prize.

GatorWings’ autonomous radio was able to navigate the various wireless obstacles developed for SC2 to thoroughly stress each team’s AI-enabled radios. GatorWings’ unique approach to the SC2 challenge helped it eke out the competition. Using an AI engine that is one-step beyond basic rule-based systems, GatorWings applied foundational reinforcement learning AI techniques to optimize each “pocket” of available spectrum.

The Spectrum Collaboration Challenge started in 2014 and since then incredible progress has been made – both in terms of what the teams can do and the sophistication of the decision engines, as well as their radio’s ability to work in much more complicated environments. This is going to push us towards making this a practical technology that can be deployed both by the military and commercial users. Through SC2’s efforts, we are moving towards using more intelligence in our radios and making things more adaptive and collaborative.

During the SC2 Championship Event, the teams’ AI-enabled radios were tested during five rounds of competitive, round-robin matches. Each round focused on a different wireless scenario with various obstacles that autonomous radios could face in the real world – from gradually shrinking bandwidth to temporal surges. At the end of each round, the two lowest scoring teams went head-to-head to determine who would move on to the next round and who was eliminated from the competition. The final five teams moved onto one last round of competitive play. This final round introduced new obstacles for the teams’ technologies to overcome, including legacy radio systems that are sensitive to interference. The teams received prize points based on their rank at the end of each round. Those points were then totaled to determine the final rankings and the first, second, and third place winners.

Throughout the event, the virtual wireless world came to life with emotional reactions from the 10 finalist teams, live commentary from the event emcees – which included Grant Imahara from the Discovery series MythBusters and Netflix series White Rabbit Project, DARPA program manager Paul Tilghman, and GNU Radio president Ben Hilburn – as well as a custom-built visualizer that showed how each radio fared at autonomously navigating the spectrum while avoiding interference with their competitive counterparts.

While GatorWings took home the top spot, the second and third place finishers were MarmotE and Zylinium, respectively. MarmotE, a team of current and former Vanderbilt researchers, took home the $1 million second place prize, while the third place prize of $750,000 went to Zylinium, a three-person start-up with expertise in software-defined radios (SDRs) and AI. Andersons, a two-person team of hobbyists and SDR enthusiasts that also successfully competed in DARPA’s 2014 Spectrum Challenge, and Erebus, a three-person company created specifically to tackle SC2, rounded out the top five.

It was truly a battle right until the end, with GatorWings beating out MarmotE by just one point. GatorWings really stepped up their design in the final year, they previously finished fifth and sixth in the preliminary tournaments. Each team took a slightly different approach to the final scenarios – some used AI to navigate the wireless spectrum like a driverless car, while others used machine learning to promote competing or collaborating solutions. In the end, the three highest ranked teams were able to maximize their use of the spectrum by skillfully collaborating with their competitors’ radios while successfully completing as many data transfers as possible.

The SC2 Championship Event provided further proof that collaborative, autonomous wireless networks are capable of beating the status quo of static, human-driven spectrum allocation. While still in the early days for exploration, these technologies show significant promise towards easing the strain increasingly placed on our wireless resources as more commercial and defense technologies become wirelessly enabled.