Smart Noise Jamming Power Adjustment Using Exploratory Deep Deterministic Policy Gradient

Smart noise jamming is an emerging barrage jam-ming and plays an essential role in electronic countermeasure (ECM). To improve the interference effectiveness of smart noise, in this paper, an exploratory deep deterministic policy gradient (EDDPG) algorithm is proposed to continuously adjust the multi-step jamming power. Firstly, multi-step jamming power adjustment is modeled as a Markov decision process (MDP). Subsequently, average jamming-to-signal ratio (JSR) at multi-functional radar (MFR) receiver is chosen as the evaluation indicator to assess the performance of multi-step noise. Moreover, the principle of noise jamming power allocation is analyzed, and a reinforcement learning framework is developed to continuously adjust multi-step jamming power. Finally, numerical results are provided to verify the validity of the proposed method.