Low SNR FM signal preprocessing method based on low-order cyclic statistics and WVD distribution

In order to solve the problem of poor noise suppression effect and long processing time of FM signal under the condition of low SNR by traditional preprocessing methods, a signal preprocessing method based on low-order cyclic statistics and WVD distribution is proposed. The method firstly calculates the cyclic mean value of incoming wave signal, and then transforms the incoming wave signal domain through IFFT algorithm with variable points. Finally, the time-frequency ridge of the FM signal is obtained by WVD time-frequency analysis, and the binary image transformation is carried out. Through simulation analysis, the time-frequency preprocessing method proposed in this paper can greatly reduce the calculation amount and improve the real-time performance while retaining the accuracy of time-frequency features, and provide effective support for subsequent sorting identification and parameter estimation.