Benchmarking Machine Learning Methods for Distributed Acoustic Sensing
Shuaikai Shi, Qijun Zong
TL;DR
This work addresses the problem of robust DAS signal classification by benchmarking a spectrum of machine learning methods, from conventional algorithms to deep neural networks, on a dataset derived from fiber-optic DAS along the Albula River. The authors employ a data-processing pipeline that transforms time-series DAS data into image-like representations and uses Cross-Spectral Density Matrices, augmented with a VQ-VAE-based representation learning stage, to improve discriminative power. Experimental results show deep learning models, particularly ResNet and Transformer-based architectures, consistently outperform traditional models such as LR and SVM, achieving maximum accuracies around 0.958 and strong F1-scores, while classical methods remain competitive in some settings. The study provides practical guidance for selecting DAS classifiers in real-world deployments and highlights future opportunities in transfer learning, domain adaptation, and interpretable, privacy-preserving approaches for scalable, robust DAS analytics.
Abstract
Distributed acoustic sensing (DAS) technology represents an innovative fiber-optic-based sensing methodology that enables real-time acoustic signal monitoring through the detection of minute perturbations along optical fibers. This sensing approach offers compelling advantages, including extensive measurement ranges, exceptional spatial resolution, and an expansive dynamic measurement spectrum. The integration of machine learning (ML) paradigms presents transformative potential for DAS technology, encompassing critical domains such as data augmentation, sophisticated preprocessing techniques, and advanced acoustic event classification and recognition. By leveraging ML algorithms, DAS systems can transition from traditional data processing methodologies to more automated and intelligent analytical frameworks. The computational intelligence afforded by ML-enhanced DAS technologies facilitates unprecedented monitoring capabilities across diverse critical infrastructure sectors. Particularly noteworthy are the technology's applications in transportation infrastructure, energy management systems, and Natural disaster monitoring frameworks, where the precision of data acquisition and the reliability of intelligent decision-making mechanisms are paramount. This research critically examines the comparative performance characteristics of classical machine learning methodologies and state-of-the-art deep learning models in the context of DAS data recognition and interpretation, offering comprehensive insights into the evolving landscape of intelligent sensing technologies.