DAS signal denoising using adaptive beamforming

Fiber-optic distributed acoustic sensing (DAS) technology has been extensively applied in many different fields, while enhancement of its signal-to-noise ratio (SNR) is always of the first priority to let its high sensitive perception ability be brought into full play. In this paper, a novel DAS signal denoising methods is proposed by utilizing the adaptive beamforming (ABF) of its array signal rather than a single point signal, for the first time. Three ABF algorithms are comparatively studied, including minimum variance (MV), eigenspace-based minimum-variance (ESBMV), and coherence factor (CF) filtering. The experimental results show that these three algorithms improve the DAS signals by a similar level of 10.6 dB, 10.6 dB, and 10.2 dB, respectively for noisy DAS signal with SNR of 29.2 dB. The processing time for the three methods is also compared, and it shows that the MV takes the shortest time of only 11 ms, which is the most promising ABF denoising method for DAS. It is highly anticipated that this ABF method could be used in high-performance DAS systems for applications in oil/gas exploration, seismic surveillance, pipeline monitoring, and submarine acoustic detection, et al.