Radar Signal Recognition under Impact Noise Based on Convolutional Neural Network

Abstract

To solve the problem of radar signal recognition under the impact noise along with the conventional Gaussian white noise, we propose a method for radar signal recognition based on Choi-Williams Distribution (CWD) time-frequency transform and convolutional neural network. The α distribution is employed to model the impact noise in radar signals. The proposed method firstly performs CWD time-frequency analysis on the radar signal. Then, two-dimensional time-frequency images obtained by time-frequency transform are fed to a lightweight convolutional neural network for deep feature extraction. Finally, a softmax classifier is used to classify and recognize the radar signals. The simulation results show that the proposed method performs well in the signal classification task, and the lightweight convolutional neural network model provides convenience for realizing FPGA hardware acceleration.