Energy-balanced clustering with Master/Slave method for wireless sensor networks

Abstract

Traditional clustering algorithms for WSN (wireless sensor networks) select only one CH (cluster head) in each cluster, which consumes energy at the CH fast and shortens the network lifetime greatly. In this paper, EBCMS (energy-balanced clustering algorithm with master/slave method) is proposed to solve this problem. The key idea of the algorithm is that one master CH and two slave CHs are chosen in each cluster. EBCMS selects the nodes with more residual energy as master CHs and slave CHs. In procedure of data forwarding, EBCMS selects those nodes as relay nodes, which have minimum energy consumption for forwarding. The simulation results prove that EBCMS improves by 86% compared with LEACH and 43% compared with RRCH in the dead time of half of nodes in the area of 500 m * 500 m, while it improves by 34% compared with LEACH and 37% compared with RRCH (round-robin cluster head method) in the area of 1000 m * 1000 m.