Clustering algorithms for wireless sensor networks using spatial data correlation

In densely deployed wireless sensor networks, spatial data correlations are introduced by the observations of multiple spatially proximal sensor nodes on a same phenomenon or event. These correlations bring significant potential advantages for the development of efficient strategies for reducing energy consumption. In this paper, spatial data correlations are exploited to group sensor nodes into clusters of high data aggregation efficiency. We define the problem of selecting the set of cluster heads as the weighted connected dominating set problem. Then we develop a set of centralized and distributed algorithms to select the cluster heads. Simulation results demonstrate the effectiveness and efficiency of the designed algorithms.