Study of weak high-frequency CW signal detection based on stochastic resonance

Abstract

CW telegraph has been largely used for high-frequency tactical communication. But there exists serious decline in high-frequency channel, which presents great difficulties to the detection and identification of CW telegraph signals. The traditional detection method based on time-frequency transformation has a poor performance in strong noise environment. An advanced detection method that combines non-linear bi-stable stochastic resonance model and time frequency analysis is proposed in this paper. Firstly, the received CW signal was modulated to a certain low frequency, and filtered by stochastic resonance model to restore its waveform from strong noise environment. Secondly, the WVD transformation and Hough transformation were utilised to extract its time domain features of the restored waveform. Thus, the detection of CW telegraph signal was easily implemented by detecting the peak values in parameters space. The simulation studies and results showed that the new method performs very well in the detection of CW signal with low signal noise ratio.