An improved Newton algorithm based on finite element analysis for extracting the Brillouin scattering spectrum features

In the distributed optical fiber sensing system based on Brillouin scattering, the photodetector offers the low signal-to-noise ratio (SNR) and can hardly demodulate complicated information. An improved Newton algorithm based on finite element analysis (FEA-Newton algorithm) is proposed to extract the estimated parameters of Brillouin scattering spectrum from noisy signals accurately. By using the FEA-Newton algorithm, the relationship of the linewidth vs center frequency shift error and SNR vs center frequency shift error are both analyzed and the Brillouin center frequency shifts are extracted. Within a certain range of initial value, the accuracy of the Newton and FEA-Newton algorithms are compared. The measured Brillouin shift of FEA-Newton algorithm is match with theoretical values. The FEA-Newton algorithm is more preferably to extract the Brillouin scattering spectrum features.