Radar Signal Sorting and Individual Recognition based on LSTM

In order to solve the problem of low signal sorting and recognition accuracy when the radiation source loses high pulse rate, this paper proposes a kind of radar signal sorting and individual recognition technology based on LSTM. First, the timing prediction performance of LSTM is used to predict the PDW parameters of the arriving pulse, and other radiation source signals are eliminated according to the threshold to complete the sorting of the target radiation source signals. The experimental results show that the lost pulse rate is less than 30% and the sorting accuracy is more than 87%. LSTM_FCNN network is used to identify the individual radiation source of the sorted signals. The simulation results show that the recognition accuracy of seven different radiation sources is more than 90% when the pulse loss rate is less than 40%.