Distributed multidimensional clustering based on spatial correlation in wireless sensor networks

摘要

Wireless sensor networks (WSNs) are used to collect various data in environment monitoring applications. The most comprehensive way of data collection is to make every sensor node report periodically its sensing data to a base node. Then the key challenge is to reduce the energy consumption due to excessive communication. The spatial clustering may save energy by partitioning the network into a set of spatial clusters with similar sensing data. For each cluster, only a few sensor nodes (samplers) report their sensing data to the base node. The base node may predict the missed data of non-samplers using the spatial correlation between sensor nodes. In this paper, we consider WSNs for complex monitoring applications where each sensor node is equipped with several sensors and thus a sensing data consists of multiple attributes. We first introduce problems of previous spatial clustering algorithms which considered single-attribute data only. Then we propose an energy-efficient distributed multidimensional clustering algorithm. Finally, we present experiment results, which indicate a reduction in data collection overhead up to 51% when compared to previous algorithms.