Dynamic Multi Level Hierarchical Clustering Approach For Sensor Networks

Wireless sensor networks have the problem of lifetime and scalability. To increase lifetime and scalability it is necessary to have control over topology of the network even when dynamically changes are observed. Dynamic clustering with adaptive feature is the best way to achieve the above. In this paper we propose a dynamic multi-level hierarchal clustering (DMH) approach for sensor networks. The proposed approach will create a dynamic system which can vary topology architecture according to traffic patterns in the active network. This approach can decide size of cluster, nodes in a cluster and level of hierarchy of a cluster and will vary according to the instant state of the system. In this approach for clustering we use nodes having multiple energy levels for energy efficient clustering along with mutual negotiation between nodes of Subsequently the cluster heads are selected periodically based on different attributes such as residual energy, node degree etc. in the proposed approach we have used mutual negotiation between nodes of the cluster; as a criteria for cluster formation, which was not used in the previously available approaches for cluster formation. Also here we have used dynamic adaptive level of hierarchy according to the traffic patterns and use the highest level of hierarchy for routing of aggregated data to the base station in the sensor network. The proposed method of clustering has given new dimension in clustering and performance of sensor network has improved drastically, especially with the increment of lifetime for network under test. The new method of clustering is also helpful in managing the sensor network with improving traffic rate and life.