Positioning Error Reduction Technique Using Spectrum Reshaping for Distributed Fiber Interferometric Vibration Sensor

Abstract

A novel positioning error reduction technique is proposed for dual Mach-Zehnder fiber interferometric distributed vibration sensor. The 3 dB bandwidth of the power spectrum of interference signal is broadened to reduce the mean square error (MSE) of the cross-correlation based positioning algorithm. A high-pass filter (HPF) whose cutoff frequency is larger than the upper frequency of signal 3 dB bandwidth is applied to reshape the original power spectrum by attenuating the magnitude of low-frequency mainlobe to the same order as that of the small high-frequency components. The usage of HPF can cause bandwidth broadening and signal to noise ratio (SNR) reduction, the influence of both effects on ultimate positioning MSE are analyzed in detail. Theoretical analysis shows that there exists a valid region for the cutoff frequency of the HPF within which the contribution of bandwidth broadening dominates the process. The technique is also experimentally verified and analyzed by field test on an installed submarine cable between two islands of Zhejiang Province, China. Field test results coincide with the theoretical predictions and show that a maximum of 7 dB reduction of positioning MSE can be achieved when optimal cutoff frequency of HPF is used.