Abstract: In this paper the proposed work is to compare the properties of the nodes in the cluster on the basis of probability of their being retrieved by other cluster or network. All this nodes and clusters contains battery level, memory, reverse signal strength indication, multicast routing information cost , time to live new node and time to live cluster on the basis of expectation clustering and maximization clustering in wireless sensor networks.

Keywords: Expectation, maximization, cluster head, energy efficiency, k-means, data aggregation, WSN.