An Improved Distributed Cluster Organization Algorithm of Wireless Sensor Networks

Dynamic cluster organization is a hotspot research field for wireless sensor networks. A new algorithm of building and regroup clusters - improved distributed cluster organization algorithm (IDCOA) is proposed in this paper. By IDCOA, nodes are clustered via the way that some nodes propose invitations, and the other sensors respond. Member nodes in the networks save the cluster´s id, but header nodes do not have to save its member´s id. The regroup of clusters is considered also. On one hand, the exiting nodes don´t affect on the other nodes; on the other hand, the new nodes can join in the network effectively. To describe the IDCOA algorithm, six time-related parameters are defined to be used for timer control; seven message types are defined to be used for nodes communication; and ten steps are defined as the implement of the IDCOA process. The simulate result indicates that IDCOA algorithm has well performance on the clustered rate and the number of cluster headers for wireless sensor networks.