A Threat Evaluation and Jamming Allocation System Against STAP Radar Based on Double Q-Learning and Chaos Genetic Algorithm

In modern warfare, the electromagnetic warfare environment is complex and extensive. To enhance operational effectiveness, various threat evaluation and jamming allocation (TEJA) systems have been proposed to provide jamming strategies for cooperative jamming platforms. However, current TEJA perform poorly when confronting Space-Time Adaptive Processing (STAP) radar. This paper proposes STAP-COTEJA, a distributed jamming evaluation and decision-making system against STAP radar. The system takes into account various radar information for threat evaluate and considers the interactions of data exchange between radars, making the result of threat evaluation more realistic. When estimating the jamming effect, the system directly simulates STAP process of radar, resulting in higher accuracy. This system also employs RL-CGA, an algorithm combining reinforcement learning and chaos genetic algorithm, to acquire optimal jamming strategy. The system is tested under a complex scenario where 7 jammers need to jam 3 STAP radars. Test results show that after using RL-CGA, STAP-COTEJA is able to obtain the optimal jamming strategy, significantly reducing radar threats. Its running time is significantly reduced compared to CGA. The better jamming effectiveness and faster running speed indicate that STAP-COTEJA can effectively counter STAP radar systems, improving the survival ability of friendly units.