Noise reduction by Brillouin spectrum reassembly in Brillouin optical time domain sensors

Abstract

In this paper, we present a novel image-processing based spectrum reassembly method to handle the noisy Brillouin gain or loss spectrum in Brillouin optical time domain sensors. The principle of the proposed Brillouin-spectrum-reassembly (BSR) Filter, is firstly dividing the Brillouin spectrum along the fiber into several segments according to the differences of Brillouin frequency shift (BFS), then shifting the Brillouin spectrum to reassemble a new image with equal BFSs and filtering the image by a conventional filter, finally shifting the Brillouin spectrum of each segment in the opposite direction to obtain the filtered Brillouin spectrum. Since the size of the filter windows in the reassembled image can be reduced greatly, the total processing time is decreased effectively. Furthermore, the BFS measurement accuracy and spatial resolution can be enhanced due to the edge-preserving image filtering strategy of the method. The measurement performance and processing speed improvements are verified by simulation and experiment. Utilizing a Brillouin optical time domain reflectometry (BOTDR) sensor, we achieve a BFS measurement accuracy improvement of 13%, a spatial resolution improvement of 46% and a processing time reduce of 35% in a 5.2 km-length G657 fiber by use of the proposed method, as compared with an excellent spectrum non-reassembly filtering method.