Automatic Parameter Tuning with Metaheuristics of the AODV Routing Protocol for Vehicular Ad-Hoc Networks

Abstract

Communication protocol tuning can yield significant gains in energy efficiency, resource requirements, and the overall network performance, all of which is of particular importance in vehicular ad-hoc networks (VANETs). In this kind of networks, the lack of a predefined infrastructure as well as the high level of dynamism usually provoke problems such as the congestion of intermediate nodes, the appearance of jitters, and the disconnection of links. Therefore, it is crucial to make an optimal configuration of the routing protocols previously to the network deployment. In this work, we address the optimal automatic parameter tuning of a well-known routing protocol: Ad Hoc On Demand Distance Vector (AODV). For this task, we have used and compared five optimization techniques: PSO, DE, GA, ES, and SA. For our tests, a urban VANET scenario has been defined by following realistic mobility and data flow models. The experiments reveal that the produced configurations of AODV significantly improve their performance over using default parameters, as well as compared against other well-known routing protocols. Additionally, we found that PSO outperforms all the compared algorithms in efficiency and accuracy.