Deep Reinforcement Learning-Based Jamming Against Multiple Frequency Agile Radars

To tackle the ever-growing complexity of radar systems in modern electronic warfare, there is an urgent requirement for more effective and adaptable electronic countermeasure systems. In practical terms, current electronic warfare systems encounter difficulties in simultaneously combating multiple frequency agile radars (FARs). Herein, we present a intelligence jamming methodology for jamming aircraft that utilizes deep reinforcement learning (DRL). The jamming system can allocate resources intelligently by observing the radar's operating modes and adapting accordingly using DRL techniques. In the proposed approach, we first introduce a jamming system model that employs a pulse-by-pulse jamming technique, combining spot jamming and frequency sweeping jamming. Then, the temporal and energy resource allocation decisions for the jamming system when jamming against a group of radars are modeled as a Markov decision process (MDP). We utilize the deep deterministic policy gradient (DDPG) algorithm, which is well-suited for solving MDP in continuous action spaces, to train and determine the optimal strategy for the jamming system. Simulation experiments validate the effectiveness of the proposed algorithm compared to conventional jamming strategies.