An Adaptive Fuzzy System in Large Scale Mobile Ad Hoc Networks

Multicast protocols in MANETs must consider control overhead for maintenance, energy efficiency of nodes and routing trees managements to frequent changes of network topology. Now-adays Multicast protocols extended with Cluster based approach. Cluster based multicast tree formation is still research issues. The tree reconstruction of cluster-based multicast routing protocol will take place if any link of the trees has malfunction or the nodes move out of the link, therefore, its robust performance is unsatisfactory. The mobility of nodes will always increase the communication delay because of re-clustering and cluster head selections. For this issue we proposed the new scheme Adaptive Fuzzy System (AFS), its fuzzy based clustering and predicting the next cluster head (CH) based their location updates with clustered group. A new location management scheme is proposed to handle the mobility of cluster members, based on a hybrid strategy that includes location updating and location prediction. In a clustered zone predicts movement of members and CH based on Kalman filtering of previously received updates and based on location updates CH will selected. Here location managements will leads to reduce cluster head selections. We used ns2 for our AFS.We present simulation results that demonstrate a significant reduce the communication delay over the traditional cluster based MANETs deployments.