A Neural Network Method For The BFS Extraction

Authors:

Abstract
We present a method for increasing the accuracy of determining the Brillouin frequency shift in a fiber strain sensor. A neural network which using the results obtained from other methods as input data and the parameters of the system is applied. The result is an additional 10% accuracy increasing, which is significant given that existing methods have already reached their limit and work at low SNR values (about 5 dB).

This publication has 7 references indexed in Scilit: