Multiple Noise Reduction for Distributed Acoustic Sensing Data Processing through Densely Connected Residual Convolutional Networks

Abstract Distributed acoustic sensing (DAS), which utilizes the entire optical fiber as the sensing medium, provides distinct advantages of high resolution, dynamic monitoring, and resistance to high temperatures. This technology finds diverse applications in the seismic exploration, oil survey, and submarine cable monitoring industries. However, DAS signals are susceptible to various kinds of noise, such as horizontal noise, optical noise, random noise, and so on, which significantly degrade the signal-to-noise ratio (SNR), this low SNR is likely to affect some subsequent analyses, such as inversion and interpretation. These mixed noises can pose a serious challenge to noise reduction in the DAS signal. To address this issue, we have developed a supervised learning-based densely connected residual convolutional denoising network (DCRCDNet), which leverages both encoding and decoding processes to extract features and reconstruct DAS data. The encoding and decoding processes enable the network to fully extract the number of features. The design of dense connectivity and residual blocks allow the network to better extract shallow to deep features, which ultimately reconstruct our DAS signal hidden in the noise. In comparison to the traditional filtering method and other deep learning methods, DCRCDNet has great potential for attenuating strong and mixed noise and extracting hidden signals.