Effective cluster based techniques for improving networking functions of mobile ad hoc networks

Abstract

Mobile Ad-hoc Network (MANET) consists of mobile nodes interconnected by wireless multi-hop communication paths. Unlike conventional wireless networks, ad-hoc networks have no fixed network infrastructure or administrative support. In the present thesis, four cluster based techniques have been proposed for improving the performance of MANET by enhancing its key networking functions along with effective clusterhead selection. The existing cluster based routing algorithms suffer from frequent change of clusterheads, resulting in the failure of sessions and network partitions. The proposed Adaptive Weighted Cluster Based Routing (AWCBRP) algorithm amends swiftly to the topological changes and accomplishes the routing efficiently with improved packet delivery ratio and minimized end-to-end delay. A Cluster-Based Service Discovery (CBSD) algorithm has been developed to achieve efficient service identification by using reliable clusterhead. A Cluster Based Congestion Control (CBCC) algorithm has been developed to support congestion control using scalable and distributed cluster based approach. The co-operation among the cluster nodes is achieved by exchanging a small amount of control packets along with the communication paths. AWCBR adapts quickly to the topological changes and establishes the routing efficiently with 97% to 98% of Packet Delivery Ratio (PDR) for 20 to 100 nodes. It improves the packet propagation delay by picking up the efficient routes that satisfies the highest consistency along with loop free attribute from the source node to the destination node and exhibits lowest endto- end delay with 0.04 to 0.11 seconds for 40 to 100 nodes. The TC-EPM algorithm stabilizes the topology of the network by regulating the connectivity power level, so that the nodes converge quickly to a desired topology independently of the initial network state. It achieves 191 nodes to be alive out of 200 nodes for 500 seconds of simulation time. newline newline newline