Incremental histogram based anomaly detection scheme in wireless sensor networks

Many mission critical wireless sensor networks require an efficient and lightweight anomaly detection scheme to identify outliers. In this paper, we propose an incremental histogram based anomaly detection scheme in order to detect the anomaly data values within the network. It first partitions the whole network into several clusters in which the cluster members are physically adjacent and data correlated. Then, the cluster head and cluster members update histogram incrementally and compare histograms in the form of kullback-leibler divergence differentially. We show through experiments with real data that the proposed anomaly detection scheme can provide a high detection accuracy ratio and a low false alarm ratio.