Vertical distance estimation of vibration sources based on distributed optical fiber sensing

The accurate estimation of vertical distance between the external vibration source and the fiber is a prerequisite for events detection and identification. Due to the limitation of non-homogeneity of fiber material and weak signals with strong background noise, traditional distance estimation of vibration sources methods have difficulty in obtaining accurate results. To meet the needs of high-precision sensing and spatial positioning, we propose a novel vibration source distance estimation method based on multi-point collaboration combined with improved TDOA. In data preprocessing, the strong noise in vibration signals is filtered out using wavelet threshold transform. Specifically, the 3-Sigma criterion and a least-squares fitting algorithm are employed to fit the outliers. Then, we adopt the short-time energy method with two-stage adaptive threshold to measure the arrival time of the vibration source. Finally, the vertical distance is calculated from the difference of arrival times combined with the arrival azimuth. The experiment results demonstrates that the proposed method has a superior performance and potential for application, where the accuracy of results is higher than 95%.