ENERGY CONSTRAINT ANALYSIS IN APPROXIMATION ALGORITHMS FOR MANETS

Abstract

A mobile ad hoc network is a simple network configured for a temporary basis without any physical infrastructure and can be dismantled when the required goal is achieved. In this type of network, nodes are allowed to move around in the geographical area i.e., nodes are mobile. A clustering MANET consists of two phases i) Cluster Formation and ii) Cluster Maintenance, and this project report suggests a customised optimization algorithm. All these clusters are formed by moving nodes and they contain a very limited number of resources which provides the fundamentals for clustering in MANETs: 1. Dynamic Topologies – nodes are mobile and hence topologies keep on changing. 2. Limited battery – mobile nodes are battery operated and a good algorithm is required to manage that as well. 3. Limited bandwidth – capacity of wireless links is less and they are also susceptible to interference and noise. 4. Limited Security – a node is allowed to join or leave any network, so it is risky to allow connection without checking. Limited resources of CH depleted most and when a CH reaches a min threshold, it transfers its duties to another node, which is elected by CH itself. This report considers various Approximation algorithms which have been utilised to optimise the search results of various problems. A CH selection in Clustering MANETs is also a multi-objective problem which can be solved using these approaches. In this thesis, GA and PSO are applied on top of the LEACH protocol for better CH selection.