期刊名称：TELKOMNIKA (Telecommunication Computing Electronics and Control)
印刷版ISSN：2302-9293
出版年度：2015
卷号：13
期号：3
页码：894-903
DOI：10.12928/telkomnika.v13i3.2010
语种：English
出版社：Universitas Ahmad Dahlan
摘要：As the traditional anomaly detection algorithms cannot effectively identify the spatial-temporal anomaly of the wireless sensor networks (WSNs), taking the CO2 concentration collected by WSNs for example, we propose the spatial-temporal anomaly detection algorithm in wireless sensor network. First we use the 3 rules to realize the anomaly detection of the adaptive threshold. Then extract the eigenvalue (average) of sliding window to be detected, construct the spatial-temporal matrix for the relationship between neighbor nodes in the specified interval, use the fuzzy clustering method to analyze the eigenvalue of adjacent nodes in spatial-temporal correlation and classify them, and identify abnormal leakage probability according to the results of the classification. Finally, use real data to verify this algorithm and analyze the parameters selected , the results show that the algorithm is high detection rate and low false alarm rate.
其他摘要：As the traditional anomaly detection algorithms cannot effectively identify the spatial-temporal anomaly of the wireless sensor networks (WSNs), taking the CO2 concentration collected by WSNs for example, we propose the spatial-temporal anomaly detection algorithm in wireless sensor network. First we use the 3 rules to realize the anomaly detection of the adaptive threshold. Then extract the eigenvalue (average) of sliding window to be detected, construct the spatial-temporal matrix for the relationship between neighbor nodes in the specified interval, use the fuzzy clustering method to analyze the eigenvalue of adjacent nodes in spatial-temporal correlation and classify them, and identify abnormal leakage probability according to the results of the classification. Finally, use real data to verify this algorithm and analyze the parameters selected , the results show that the algorithm is high detection rate and low false alarm rate.
