A Superfast Brillouin Frequency Shift Estimation Method Based on the Mixed Moment for the Distributed Optical Fiber Sensor

To estimate Brillouin frequency shift with high accuracy and less computational burden, the mixed moment is used to estimate it for the distributed optical fiber sensor. Its computational burden is much less than that of the typical algorithms. Meanwhile, it is free of curve fitting and initialization of fitting parameters. The optimal order in the proposed algorithm is determined. The accuracy and computation time of the proposed algorithm, the typical spectrum fitting algorithms based on the pseudo-Voigt model, the quadratic polynomials, and the cross-correlation algorithm are compared for the simulated and collected Brillouin spectra with different signal-to-noise ratios. The results reveal that the computation time of the proposed algorithm is only 1/6464~1/3986, 1/33~1/24, 1/111~1/72 of that of the other three algorithms. However, its accuracy is close to that of the spectrum fitting algorithm based on the pseudo-Voigt profile and is a little higher than that of the other two algorithms