Raytheon demonstrates AI-enabled radar warning receiver

Raytheon has completed flight testing of an AI-enabled radar warning receiver for 4th generation aircraft, marking the first time that this has been done, according to a 24th February company press release. The capability was integrated into an existing radar warning receiver, improving the system’s ability to sense and understand threats. 

The upgrade consists of the Cognitive Algorithm Deployment System (CADS) which combines an “embedded graphics processing unit” with a computing stack from Deepwave Digital, a US company that specialises in building edge computing systems for radio frequency applications. The demonstration appears to be an example of software defined defence, where legacy systems are improved through software and computing upgrades, rather than extensive research and development to build a brand new sensor. 

“CADS performed successfully during flight testing and demonstrations on an F-16 at the Air National Guard’s test range near Tucson in December,” the press release states. It adds that the tests incorporated containerised AI/ML (artificial intelligence/machine learning) techniques from the Georgia Tech Research Institute, Vadum, Inc., and Raytheon’s own cognitive electronic warfare team. 

A containerised AI/ML application is where everything needed to run an algorithm is integrated into a single software package. This includes the code, runtime, system tools, system libraries, and settings. A container allows the algorithms to be deployed to the computer, be that on a cloud computer or physical one, and be fed with real-world (as opposed to training) data. The container then runs the models that are built into it and returns a result to the user through its software interface. The container also allows the models to be moved between computers and rapidly scaled as required. 

“CADS’ ability to quickly process data and run third-party algorithms that prioritize threats, with almost no latency will significantly enhance survivability for military personnel,” Bryan Rosselli, president of Advanced Products and Solutions at Raytheon said. 

The CADS, including its cognitive radar capability, was initially tested on Raytheon’s flight test aircraft before being integrated onto the F-16. The demonstration is presumably designed to prove the capability and represent it to F-16 users that may be considering a radar warning receiver upgrade. BAE Systems is delivering the AN/ALR-56M radar warning receiver for US Air Force F-16s and C-130Js according to a 2024 announcement. The profile for the system on the BAE website does not indicate any specific AI capabilities, but does indicate advanced filtering and signal isolation. It stands to reason that the Raytheon upgrade may not find its first customer in the US. 

Raytheon is not the first company to pursue AI on the edge for electronic warfare (EW). Helsing and Saab announced a partnership in 2023 to provide AI-enabled EW for a small number of Eurofighter’s in Germany’s fleet which leveraged existing hardware and Helsing’s software and AI. 

Calibre comment

Cognitive electronic warfare (EW) is becoming more important as western forces seek to improve the survivability of their air forces against adversary air defence capabilities. There are two key aspects to cognitive EW, which is underpinned by AI: The first is to understand the threat, as the Raytheon announcement indicates. In this application, the software and AI works to understand small nuances in the signals received by the radar warning receiver and provide a better indication and advice to the pilot. It might, for example, better understand what a threat radar is doing than legacy systems, which would simply indicate that there is a threat radar. The second aspect is delivering electronic jamming against a target. The typical approach can be generalised as one of brute force; if your jammer can generate more noise than the adversary jammer, then it will succeed until that is not the case. Cognitive EW envisages the use of AI to deliver more nuanced and bespoke effects, spoofing an adversary’s radar making it think that there are no threats, for example. Together, these types of capabilities are hoped to reduce the efficacy of layered air defence networks as developed by Russia, China, and to some extent, Iran, improving the survivability of western strike packages. 

By Sam Cranny-Evans, published on 25th February 2025.