Drift-Robust Differential Processing for Improved Measurement Fidelity and Stability in Distributed Acoustic Sensing

Abstract

Measurement drift, reference instability, and frequency-dependent sensitivity owing to traditional temporal differencing limit the performance of distributed acoustic sensing (DAS). Moving differential (MD) and fixed differential (FD) operators function as implicit measurement operators. These operators exhibit comb-filter-like responses, which result in systematic sensitivity loss under specific conditions. This study proposes a hybrid differential (HD) operator that integrates MD and FD via adaptive short-term/long-term averaging weighting and quiet run-triggered reference update. The benefit of this approach is the maintenance of a stable measurement baseline along with short-term vibration detection without reliance on the selection of a differencing interval. Experimental validation employing direct- and coherent-detection DAS systems over fiber lengths of up to 10.3 km indicates enhanced measurement fidelity across various excitation frequencies and amplitudes. The HD method improved the average Signal-to-Noise Ratio (SNR) by 10.07 dB over MD and 9.52 dB over FD and significantly enhanced low-strain vibration detection. For example, at 0.5 Vpp, 200-Hz excitation, HD achieves 11.39-dB SNR, compared with 5.42 dB for FD and 0.51 dB for MD. It also maintains stable sensitivity across the 20-Hz–1-kHz range, indicating improved measurement stability and repeatability without increasing computational complexity.