Performance evaluation and simulation of AODV using network simulator 3

Abstract

Recent interest in the field of Plug-in Hybrid Electric Vehicles (PHEVs) becoming a wireless technology has led to tremendous revelations. If the vehicles are considered as wireless nodes of a network, there are endless possibilities of controlling traffic congestion, improving driver safety and developing alert systems in case of accidents. Such a network is known as a Vehicular Ad-hoc Network (VANETs) and in many ways very similar to the well know Mobile Ad-hoc Network (MANETs).However, there are differences since VANETs entail many restrictions for the nodes. An interesting area of research is to see how routing protocols functions in both these networks.

This Final Year Project primarily focuses on understanding the functioning of the Ad-hoc On Demand Routing Protocol under different scenarios created with simulations using ns-3 such as effect of mobility, network size etc.A generic comparison is done to some other routing protocols as well.

One aspect that requires great understanding is that PHEVs run on a limited battery life and it is not feasible in the case of a large network to keep recharging individual nodes as and when they start running out of battery. Hence, it is imperative that AODV focuses on conserving the energy of the network as long as possible. Conventional AODV is energy unaware since its routing decisions have no parameters taking the battery lives of PHEVs into consideration. In this case, if a node is being used repetitively, it will run out of battery and soon switch off resulting in a link breakage. Many link breakages will result in partitioning of the network. An Energy Efficient AODV is discussed in this project which does keep the “cost of battery life” in mind while initiating the route discovery process.

After analysing the performance of AODV under several scenarios, some optimisation approaches have been suggested in order to improve the accuracy and efficiency of the protocol.