DARPA has selected BAE Systems for a three-year program to develop new artificial intelligence algorithms for rapidly deciphering unique radio frequency signals, the company said on Tuesday.

BAE Systems received a $9.2 million deal for the Radio Frequency Machine Learning System (RFMLS) where it’s tasked with building on existing machine learning work to find new applications to be applied toward improving detection of adversaries’ jamming and signal disruption attacks.iStock Cyber Lock

“The inability to uniquely identify signals in an environment creates operational risk due to the lack of situational awareness, inability to target threats, and vulnerability of communications to malicious attack,” John Hogan, director for BAE Systems’ Sensor Processing and Exploitation product line, said in a statement. “Our goal for the RFMLS program is to create algorithms that will enable a whole new level of understanding of the RF spectrum so users can identify and react to any signals that could be putting them in harm’s way.”

Scott Kuzdeba, BAE Systems’ RFMLS principal investigator, told Defense Daily work with DARPA started in June and will occur over three phrases. Phase one is focused on developing new approaches for applying machine learning to the radio frequency domain.

“BAE Systems has extensive prior experience applying machine learning to RF signals and will be leveraging knowledge from our background in machine learning and artificial intelligence research.  Within our autonomy technology portfolio, we previously explored machine learning within RF applications under the DARPA Communications Under Extreme RF Spectrum Conditions and Adaptive Radar Countermeasures programs,” Kuzdeba said.

Phase one will conclude next year with independent government testing of the new AI algorithms. The second and third phase will also each last a year.

BAE System researchers’ are also tasked with applying the algorithms to differentiate between important and irrelevant radio signals in real time.

“Modern data-driven machine learning research has enabled revolutionary advances in image and speech recognition and autonomous vehicles. At a time when adversaries have built capabilities to disrupt the RF spectrum, it has become critical to explore how machine learning could be applied to traditional RF signal processing,” company officials wrote in a statement.