A New Method for Polyphase Coded Radar Signals Recognition Based on Dual Channel Deep Learning

This paper presents a new method for recognizing polyphase coded radar signals based on dual-channel deep learning. The proposed method includes two stages. In the first step, the Morse Wavelet and continuous Wavelet transforms are used to extract signal features in the time and time-frequency domain. In the second, dual-channel deep learning is designed to recognize radar signals. The first channel uses the extracted feature in the time-frequency domain, and the second uses the phase change of the signals in the time domain. The proposed method’s performance is evaluated with simulated signals in a MATLAB environment. The simulation results approved that the proposed method provides a higher accuracy recognition than other methods, which only use single-channel deep learning with signal feature extraction in the time-frequency domain.