Individual Recognition of Big Data Radar Digital Waveform Based on Long Short-Term Memory Network

Individual recognition of Big Data radar digital waveform is based on the classification of target individuals using radar target waveform. Traditional recognition methods suffer from problems such as low accuracy and complex recognition processes. To address these issues, this paper proposes an individual recognition method for radar digital waveform based on Long Short-Term Memory (LSTM) network. The actual collected radar target digital waveform is simulated to generate big data of radar digital waveform, and then the basic frequency domain feature data is constructed. The LSTM network is used to extract the individual data features of the radar, and the network parameters are iteratively trained using the SGD optimization algorithm to achieve effective recognition of individual radar signals. Experimental results show that the radar signal individual recognition method based on LSTM has significantly improved accuracy compared to SVM and XGBOOST methods.