Radar signal recognition based on time-frequency feature extraction and convolutional neural network

To solve the problem of difficult feature extraction and low recognition rate of radar signal under low signal-to-noise ratio, this paper proposes a radar signal recognition method based on time-frequency feature extraction and convolutional neural network. This method uses short-term Fourier transform (STFT) to obtain two-dimensional time-frequency images of radar signals, and then sends the images to convolutional neural networks for deep feature extraction, and realizes the classification and recognition of radar signals through convolutional neural network classifiers. The simulation results show that for different intra-pulse modulated radar signals, when the signal-to-noise ratio is -5dB, the overall recognition accuracy of the proposed model can reach more than 93%, which effectively solves the problem of low radar signal recognition rate under low signal-to-noise ratio.