Abstract:

Because of the extremely low signal-to-noise ratio (SNR) of pulsar signals, it is difficult for the traditional denoise algorithms to suppress noise while preserving details such as micro-pulses. Therefore a new denoising algorithm which is based on the improved wavelet spatial correlation filtering is proposed. Firstly, the defects of the wavelet threshold and spatial correlation denoising algorithms in the extremely weak pulsar signals’ processing are analysed. Then, the wavelet spatial correlation filtering algorithm is improved according to the characteristics of pulsar signals, such as strict periodicity, correlation of adjacent periods, low SNR, radiation in the window and so on. Finally, the experiments are made with the pulsar signal observed by the Parks observatory and Rossi X-ray timing explorer. The results show that the proposed algorithm can preserve more details and provide a notable improvement in SNR.