Issue No. 06 - June (2011 vol. 60)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TC.2010.227
Samer Samarah , Yarkouk University, Jordan and University of Ottawa, Ottawa
Azzedine Boukerche , University of Ottawa, Ottawa
Alexander Shema Habyalimana , University of Ottawa, Ottawa
Recently, Knowledge Discovery Process has proven to be a promising tool for extracting the behavioral patterns of sensor nodes, from wireless sensor networks. In this paper, we propose a new kind of behavioral pattern, named Target-based Association Rules (TARs). TARs aim to discover the correlation among a set of targets monitored by a wireless sensor network at a border area. The major application of the Target-based Rules is to predict the location (target) of a missed reported event. Different data preparation mechanisms for accumulating the data needed for extracting TARs have been proposed. We refer to these mechanisms as Al-Node, Schedule-Buffer, and Fused-Schedule-Buffer. Several experiment studies have been conducted to evaluate the performance of the three proposed data preparation mechanisms. Results show that the Fused-Schedule-Buffer scheme outperforms the selected schemes in terms of energy consumption.
Wireless sensor networks, behavioral patterns, data mining.
S. Samarah, A. Boukerche and A. S. Habyalimana, "Target Association Rules: A New Behavioral Patterns for Point of Coverage Wireless Sensor Networks," in IEEE Transactions on Computers, vol. 60, no. , pp. 879-889, 2010.