Clustser based probabilistic approach to improve MANET performance

Clustering is adapted in Mobile network to provide controlled communication during broadcast communication. For Security, Mobility, Architecture Adaption and Energy Adaptive Mobile Networks, Cluster based communication is commonly a challenge. In this paper, a probabilistic based analytic model is presented to improve the clustering. The work is here presented as a layered model. In first stage of this model, centralized cluster head election is performed based on energy, distance, probability adaptive analysis. In second layer, the members with in cluster are identified based on communication parameters analysis and the probabilistic load destination. In this layer, the parameters: density, packet loss rate, delays and hop count is used as the cluster member selection. In final stage, the cluster adaptive controlled communication is performed over the network. The presented work is implemented in Opnet environment. The results show that the work has increases the communication rate and minimized the communication loss and communication delay.