A Modified Cluster Head Selection Algorithm Based on Random Waiting
Xia Haiyan, Xia Haiying · 2008
Cluster head selection in LEACH is characterized with its distribution algorithm of high randomicity, which makes the distance between the cluster heads unevenly. This causes great energy cost both in setup and data transmission phase. To solve the problem mentioned above, a modified cluster head selection algorithm based on random waiting is proposed for the wireless sensor network composing of trivial sensor nodes, which divide the setup phase into three sub-phases including cluster head pre-selection, cluster head campaign and cluster head announcement. Based on the random waiting cluster head advertisement message, the cluster head is selected dynamically among the candidates within an optimum local area. This will make the location and the number of the cluster heads more reasonable. Simulations show that cluster heads selected by the algorithm distribute more evenly and the energy cost needed in setup phase is reduced exponentially within a large-scale network.