SCR: A Sub-clustering Approach for Redundancy Reduction Based on Data Correlation in WSNs

Redundancy in wireless sensor networks plays an important role in analysis, monitoring and decision making. However, in dense networks where nodes are heavily deployed in an area of interest, the transmission of large amount of redundant data can be a critical issue for energy-constrained wireless sensor networks, as it leads to energy depletion for the whole network and shortens its lifetime. Typically, battery-operated sensor nodes are often placed in difficult-to reach places, so re-charging or changing the batteries is costly. Clustering-based data aggregation approaches are adopted to deal with redundancy. In this paper, we present a sub-clustering approach (denoted SCR) to reduce data redundancy and to save energy at the level of clusters. It exploits the data correlation between nodes and defines virtual neighborhoods based on the readings and physical distances. One representative node in each sub-cluster communicates with the cluster head. Our work reduces significantly redundancy and energy consumption.