Fast and accurate Brillouin frequency shift extraction in Brillouin optical time domain reflectometry (BOTDR) distributed fiber sensor by using ensemble machine learning algorithm

A S K Almoosa,M S D Zan,M F Ibrahim,Y Tanaka,A E Hamzah,N Arsad
DOI: https://doi.org/10.1088/1742-6596/2411/1/012012
2022-12-24
Journal of Physics: Conference Series
Abstract:To improve the Brillouin frequency shift (BFS) resolution measurement and processing time of the differential cross-spectrum Brillouin optical time domain reflectometry (DCS-BOTDR) fiber sensor, our team suggests employing the ensemble machine learning (EML) technique. Because it gave the best BFS resolution compared to the other T L cases, we used the BFS distribution data recorded by the pulse duration T L =14 ns case as ground truth to train the EML model in this work. After that, we tested the EML model for T L =4, 60, and 90 ns cases. We improved the BFS resolution for all T L situations by approximately 2.85 MHz, comparable to our resolution when T L was equal to 14 ns. This result demonstrates that the EML algorithm is reliable, efficient, and highly accurate in its predictive capabilities. Additionally, we have documented a rapid processing time of approximately one second. In addition, we have successfully demonstrated 20 cm spatial resolution measurement for T L =60 and 90 ns, which was not previously possible with the usual DCS-BOTDR signal processing method.
What problem does this paper attempt to address?