Density-Based Distributed Elliptical Anomaly Detection in Wireless Sensor Networks

Abstract:

Data measured and collected by WSNs is often unreliable and a big amount of anomaly data exist. Detecting these anomaly in energy-constrained situations is an important challenge in managing these types of networks. To detect anomalies induced by the decrease of battery power, we use HyCARCE algorithm but it has the problem of low detection rate and high false positive rate when the input space consists of a mixture of dense and sparse regions which make the anomalies form clusters. The paper presents a density-based algorithm to separate the normal cluster from all clusters. The performance of this algorithm is tested on a subset of the data gathered from a real sensor network deployed at the Intel Berkeley Research Laboratory in the USA and this density-based method has a better detection performance than HyCARCE algorithm.

• Periodical: Applied Mechanics and Materials (Volumes 249-250)
• Pages: 226-230
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.249-250.226
• Online since: December 2012
• Authors: Shu Hua Ma, Jin Kuan Wang, Zhi Gang Liu, Hou Yan Jiang
• Keywords: Anomaly, Clustering, Density, Detection Rate, Elliptical

References

This article has no references.