Robust method for BOTDA sensing information extraction in the Fourier transform domain.

Abstract

Most of the existing schemes for extracting the Brillouin frequency shift (BFS) are based on the line shape of the Brillouin gain spectrum (BGS) curve. However, in some circumstances, such as in this paper, there is a cyclic shift in the BGS curve, causing difficulty in obtaining the BFS accurately with traditional methods. To solve this problem, we propose a method for extracting Brillouin optical time domain analyzer sensing information in the transform domain-the fast Fourier Lorentz curve fitting method. It shows better performance especially when the cyclic start frequency is near the BGS central frequency position or when the full width at half maximum is large. The results show that our method can obtain BGS parameters more accurately in most cases than the Lorenz curve fitting method.