Markov model based circular frequency feature extraction method for electronic communication signal anti-jamming

This paper proposes a wavelet packet thresholding noise reduction algorithm for the problem of large signal frequency deviation and large timing error caused by the interference of electronic communication signals, and at the same time researches the spectral overlapping signal separation algorithm based on the cyclic smoothness of the signal. The algorithm sets up a frequency shift filter, including a frequency shift portion and a conjugate frequency shift portion at the receiving end, which facilitates signal separation using the temporal and spectral correlation of the signal. In order to solve the signal feature extraction problem, the discrete hidden Markov model structure is used for signal feature extraction by combining the characteristic intervals of the signal waveform. Simulation experiments confirm that the algorithm in this paper has a large advantage at lower signal-to-noise ratios, and when the signal-to-noise ratio is as low as −15 dB, the signal monitoring success rate of the cyclic correlation matched filter is about six times higher than that of the traditional matched filter. In the comparison experiments with SVM method and BP method, the completeness of the feature extraction results of this paper’s method reaches 95% on average, and the accuracy rate always stays above 91%, and the extracted signal features have high completeness and accuracy. It shows that the excellent performance of the algorithm designed in this paper can make the problems such as large signal frequency deviation and large timing error caused by electronic communication signal interference to be improved to some extent.