Machine Learning-Powered Anomaly Detection in Wireless Sensor Networks: A Survey
Xin Wang, Ruixue Luo, Qing 'an Li · 2024
This survey provides a comprehensive review of machine learning-based anomaly detection methods in wireless sensor networks (WSNs). WSNs are integral in various applications, ranging from military surveillance to environmental monitoring, but their open nature makes them vulnerable to security threats. Traditional rule-based anomaly detection methods often fail due to the high-dimensional and dynamic nature of WSN data. Machine learning techniques, including supervised and unsupervised learning, offer promising solutions by automatically learning from data to detect anomalies. This paper reviews the state-of-the-art in machine learning for WSN anomaly detection, discusses the strengths and limitations of different methods, and highlights the challenges of implementing these systems in resource-constrained environments.