ENERGY-BALANCING, LOCAL DATA CORRELATION-AWARE CLUSTERING ALGORITHM FOR WIRELESS SENSOR NETWORKS

Geographically proximate sensor nodes usually temporally and spatially correlated in wireless sensor networks (WSNs). Clustering is considered to eliminate data redundancy and improve in-network data aggregation efficiency. In this paper, an energy-balancing, local data correlation-aware (LDCA) clustering algorithm is proposed for WSNs. Comprehensively, considering the data correlation, energy consumption, communication distance, and other factors, we designed an average entropy and a data correlation coefficient (DCG) to make clustering and aggregation performance more effective. It not only measures data correlation properly but also reduces data volume. We also use the sensor's residual energy as one of the key elements in the cluster-head-selection phase to achieve energy balance. Simulation results indicate that the LDCA clustering algorithm achieves a higher aggregation ratio and performs better with respect to energy consumption and load balance compared to other algorithms.