Competitors from around the world came together this month for the preliminary round of DARPA’s Spectrum Collaboration Challenge (SC2) at The Johns Hopkins University Applied Physics Laboratory (APL) in Laurel, MD. This was the first event of the three-year long tournament designed to generate new wireless paradigms and access strategies in which radio networks enhanced with artificial intelligence (AI) will autonomously collaborate and reason about how to share the increasingly congested electromagnetic (EM) spectrum. Months of development and hundreds of scrimmage hours paid off for 10 teams, who each walked away from the December 13 event with $750,000 in prize money. The winning teams demonstrated software defined radio (SDR) designs that were most effective at collaborating with competing radio designs to jointly optimize spectrum usage.
As the commercial and military appetite for spectrum access continues to increase exponentially, the century-old practice of managing the EM spectrum, which relies on the manual assignment of frequencies, is being strained. To help better manage the finite spectrum resources needed to meet this growing demand, SC2 aims to create a more nimble and efficient process by applying autonomy to spectrum management.
“The concept of dynamic spectrum access isn’t new but previous attempts have focused only on isolated environments,” said Paul Tilghman, the DARPA program manager behind SC2. “With this Challenge, we are attempting to achieve autonomous spectrum management at scale. We believe the key to doing this is to have the networks actually work with each other, or collaborate, and use autonomy to determine not just the best use of the spectrum for one system, but the best use of spectrum for all systems. Our goal is to maximize the overall utility of spectrum across all different networks and wavelengths that are simultaneously operating together.”
For the preliminary event, 475 fully autonomous matches were run with the 19 qualified teams’ radio designs in SC2’s custom testbed environment, known as Colosseum. The final matches for the first event were carried out across six different communications scenarios designed to mirror real-world congested EM environments, but with more complexity than existing commercial radios are equipped to handle. The competing teams faced fluctuating bandwidths and interference from other competitors as well as DARPA designed bots that tested and challenged their radio designs. Each team’s radio performance was scored based on its collaborative spectrum sharing abilities.
At the event’s conclusion, the 10 highest scoring teams were:
- MarmotE from Vanderbilt University
- SHARE THE PIE from BAE Systems with Eigen LLC
- Zylinium from a Maryland-based startup
- Erebus, consisting of three independent engineers and software developers
- SCATTER from IDLab, an imec research group at Ghent University and University of Antwerp, and Rutgers University
- GatorWings from University of Florida
- Sprite from Northeastern University
- Strawberry Jammer from Northrop Grumman
- Optical Spectrum, consisting of two independent LIDAR engineers
- BAM! Wireless from Purdue University and Texas A&M University
“The qualifying teams demonstrated exactly the cross-section of wireless technologies we were hoping a Challenge-based program would create—from radios that divide the spectrum in frequency to those that divide in time to radios that adapt bandwidth, perform frequency hopping, spatial beam-forming, and reprioritization of wireless traffic,” said Tilghman. “Some nascent AI capabilities were also present in the competitor field. All of these designs created a microcosm representing the growing diversity in the wireless arena.”
Each of the 19 competing teams qualified for the competition by demonstrating that their radio network could successfully transmit wireless data in an interference-free environment, similar to today’s paradigm of planned spectrum usage. The competition found that when two radio networks were asked to share the spectrum, the top performing teams were successful at adapting their spectrum usage so that both networks could successfully transmit with minimal interference. Fully autonomous sharing of the spectrum with three simultaneous wireless technologies however, remains a difficult challenge. “When three different technologies attempt to coexist simultaneously there is a smaller set of overlapping strategies that will fulfill each individual radio network’s needs. This causes conflict and requires a higher degree of agility and reasoning, which will be required to be successful in the next phase,” noted Tilghman.
The next preliminary event will further challenge competitors with an interference environment beyond what existing commercial and military radios can handle—upping the number of simultaneous wireless network types from three to five, and raising the total number of radios from 15 to 50. According to Tilghman, “One way of thinking about this progression is to think of the first phase as representing squad-level interactions, the second phase moves to platoon-level interactions, building to the finale, which represents the company level.”
The top scoring team was MarmotE, which built on its experiences from the first DARPA Spectrum Challenge in 2013-2014 for this event. “These kinds of very ‘hands-on’ challenges actually force us, and also teams with very deep theoretical backgrounds, to really put their ideas into practice,” said team lead Peter Volgyesi. “That’s a very important step to make ideas work and be useful in the long term. That’s why I like this challenge—it forces professors, graduate students, and engineers to put what they already know into practice.”
The 10 winning teams each received $750,000 in prize money to help prepare for the second preliminary event, which will be held in December 2018. Over the next 12 months, they will have continued access to the Colosseum to test their designs and a chance to participate in scrimmages against other competitors. Although only 10 teams were awarded prize money, all 19 participating teams are automatically qualified to move on to the next phase, moving one step closer to competing in the finale in 2019.
“We want to find the radio design that can find a way to collaborate and share spectrum with almost any other radio. We need to make sure we keep a large and interesting number of radio designs in the competition, so in addition to the 19 teams from the first preliminary event, we’re currently looking for new submissions for the next phase of