Topology Management Using Clustering in Adhoc Network

Abstract

Efficiency of any adhoc network also depends on the management of its moving nodes. The basic idea for topology management is to elect the leader and all the moving nodes' information collected by it. This is basically referred as clustering in adhoc network. Also self organized and distributed approach is applicable for this. A highly dynamic topology is a distinguishing feature and challenge of a mobile ad hoc net- work. Links between nodes are created and broken, as the nodes move within the network. This node mobility a ects not only the source and/or destination, as in a conventional wireless network, but also intermediate nodes, due to the networks' multihop nature. The resulting routes can be extremely volatile, making successful ad hoc routing dependent on e ciently reacting to these topology changes. In order to better understand this environment, a number of characteristics have been studied concerning the links and routes that make up an ad hoc network. Several network parameters are examined, including number of nodes, network dimensions, and ra- dio transmission range, as well as mobility parameters for maximum speed and wait times. Here di erent clustering techniques are reviewed and summarized to manage the topology of the network which I have proposed as an enhancement to the highest degree clustering scheme which provides load balancing in the network by managing the history about how many times each node has served as cluster head in the network. To manage and update the topology, Cluster structure is updated after xed interval of time. Simulations are carried out using network simulator GLOMOSIM.