Hits:

Indexed by:会议论文

Date of Publication:2016-01-01

Included Journals:CPCI-S

Page Number:649-657

Key Words:WSNs; Anomaly Detection; Hypercube; KNN; Events; Spatiotemporal Correlations

Abstract:The past few years have seen an increased interest in the potential use of wireless sensor networks (WSNs) in applications. Unreliability and a dynamic nature are frequently present in the field of WSN, making anomaly detection necessary. Although events are often functions of more than one attribute and the energy in sensors is limited, the combination of data fusion and spatiotemporal correlations can overcome these limitations effectively. In this paper, we propose a spatiotemporal correlation-based anomaly detection model to differentiate normal and abnormal events in WSNs. The update phase occurs when abnormal events are detected. We demonstrate the usability and advantages of applying the spatiotemporal relevance in anomaly detection. Experimental results indicate its high performance in handling multi-dimensional sensor data.