The Improved Wavelet Denoising Scheme Based on Robust Principal Component Analysis for Distributed Fiber Acoustic Sensor

Aiming at the problems of large background noise of acoustic signals collected by distributed fiber acoustic sensor (DAS) system and unsatisfactory noise filtering effect of conventional filtering methods, in the study, an improved wavelet denoising scheme based on robust principal component analysis (RPCA) is proposed. The acoustic signal collected by DAS is first separated by the RPCA algorithm, and then the separated noisy signal is denoised by combining the improved wavelet threshold algorithm with a low-pass filter. The test results show that the proposed scheme can improve the signal-to-noise ratio (SNR) by 12–22 dB, and the perceptual evaluation of speech quality (PESQ) and correlation coefficient (CC) by 0.6–0.9 and 0.44–0.68 respectively, which can effectively improve the acoustic signal quality. The proposed scheme can accurately restore acoustic signals in noisy environments, which is of great significance in the fields of acoustic communication, recognition, and military reconnaissance.