RADAR JAMMING STATE PREDICTION METHOD BASED ON CNN-LSTM

By setting the strategy, the jammer adaptively transmits the appropriate jamming against the received radar signal, which poses a great threat to modern radar. Moreover, they also provide the potential for predicting jamming in radar systems. Radar jamming state prediction is crucial for the effectiveness and real-time performance of anti-jamming, and accurate prediction for the future state of jamming can guarantee the initiative of radar in electronic warfare. In order to better explore the effective information contained in the radar countermeasure data, a radar jamming state prediction method based on Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) is proposed for the characteristics of nonlinearity and time series of jamming state data. By incorporating the received jamming characteristics, including jamming type and parameters, along with the current radar status information as inputs into the network, the prediction of future jamming states can be achieved. Multiple comparative simulations have demonstrated the effectiveness of the proposed method, as well as its advantage in terms of time-consuming tasks.