Machine learning methods for identification and classification of events in ϕ-OTDR systems: a review

Abstract

The phase sensitive optical time-domain reflectometer ( φ -OTDR), or in some applications called distributed acoustic sensing (DAS), has been a popularly used technology for long-distance monitoring of vibrational signals in recent years. Since φ -OTDR systems usually operate in complicated and dynamic environments, there have been multiple intrusion event signals and also numerous noise interferences, which have been a major stumbling block toward the system’s efficiency and effectiveness. Many studies have proposed different techniques to mitigate this problem mainly in φ -OTDR setup upgrades and improvements in data processing techniques. Most recently, machine learning methods for event classifications in order to help identify and categorize intrusion events have become the heated spot. In this paper, we provide a review of recent technologies from conventional machine learning algorithms to deep neural networks for event classifications aimed at increasing the recognition/classification accuracy and reducing nuisance alarm rates (NARs) in φ -OTDR systems. We present a comparative analysis of the current classification methods and then evaluate their performance in terms of classification accuracy, NAR, precision, recall, identification time, and other parameters.

Distributed fiber optic acoustic sensing system intrusion full event recognition based on 1-D MFEWnet

“…An important area of artificial intelligence is machine learning [9], and the accuracy of intrusion event recognition has greatly increased in recent years with the introduction of machine learning technologies. The underlying idea is the extraction of signal features, which is then integrated with traditional classifiers to classify intrusion signals [10][11][12][13].…”

A recognition method of less databases by using decoupling parallel CNN based on multiple feature with DAS

“…Overall, the combination of DAS technology with AI represents a powerful approach to real-time and accurate perimeter security monitoring. By leveraging the sensitivity and accuracy of DAS technology with the pattern recognition capabilities of deep learning, it is possible to efficiently detect and classify different types of intrusion events along the entire length of the fiber optic cable [13][14][15][16]. Yang et al, proposed a novel real-time action recognition method for longdistance pipeline safety early warning systems based on a coherent Rayleigh scattering distributed optical fiber sensor under various hardware conditions with accuracy of 99.26% (500 Hz) and 97.20% (100 Hz) [17].…”

Explainable feature selection by SHAP for signal recognition of Ф-OTDR system

“…With the improvement of sensing range and spatial resolution, the Ф-OTDR system can be applied in more and more complex environments with various interference factors, which cause the degradation of signal recognition ability. To tackle with this issue, datadriven methods based on classification algorithms of machine learning (ML) have been applied in Φ-OTDR signal recognition [4].…”