Improved Cluster Based Routing Protocols through Optimal Cluster Head Selection for MANET

Abstract

Mobile ad hoc network (MANET) is a wireless network that communicate mobile hosts without the need of fixed infrastructure. They are characterized by mobility of mobile node (since mobile nodes in MANET move in and out) and limited resources (such as battery power). Mobility causes change in topology of network. Because, of this to create stable network is the most challenging task. It is also the same when battery power of an intermediate node is drained. To overcome these exiting problems clustering is essential, and used to control the topology of network. In clustered network most battery is consumed during communication to select cluster heads and when it routes where the data is sent from source to destination. Hence, optimal clustering algorithm has received significant research attention. A number of clustering algorithm has been proposed to improve cluster based routing protocol like weighted clustering algorithm (WCA), optimal stable clustering algorithm (OSCA), The Lowest ID clustering algorithm (LID), The highest connectivity clustering algorithm (HCC) etc. Still, they need enhancement to create stable, and optimal clustered network for mobile ad hoc network(MANET). This thesis focuses on improving CBRP(cluster based routing protocol) protocol, which is a clustered network to create stable and long-lived network. During cluster head (CH) selection, cluster head, and backup cluster head (BCH) is selected based on the parameters (residual energy or degree centrality). A node has high degree centrality or high residual energy is selected as cluster head and second highest is selected as backup cluster head. In member joining, the member is join based on parameters (relative mobility behavior, and distance of node) until it reaches its optimal member node. Mobile node which has the same relative mobility behavior and minimum distance to cluster head they grouped in the same cluster. Then there is no re clustering because there is backup cluster head and cluster maintenance, during cluster maintenance when a CH fail due to some reasons backup CH replaces CH and it finds its backup CH based on priority factor. The proposed work has been evaluated by using network simulator 2. The simulation result verify the modified cluster based routing(MCBRP) protocol is more energy efficient by selecting optimal cluster head, and by balancing the load based on energy model.