Performance Analysis of Dynamic Source Routing Protocol for Ad Hoc Networks Based on Taguchi's Method

Abstract

Mobile Ad Hoc Networks (MANET) are self-organizing, multi-hop wireless networks. The MANETs are rapidly deplorable due to the absence of a fixed infrastructure. A number of routing protocols have been proposed in solving routing problems for this type of network, to provide and maintain better connectivity in the network. Before being implemented in real life, the performance of a routing protocol is evaluated primarily through the simulation experiments. This paper presents Taguchi’s method to optimize networks parameters of dynamic source routing protocol for such mentioned networks. Performance analysis of the drop rate are conducted based on three factors namely terrain size, pause time, and node velocity. An L4 orthogonal array was used to design the experiment. The study indicated that, among the factors considered, terrain was found to have strongest effects, followed by pause time. By taking into consideration factorial effects, the optimal levels were found to be A1, B1, and C1, corresponding to terrain size of 25 × 25 m2, pause time of 15 seconds, and node velocity of 0.72 m/s. Using these factor-levels combination, a minimum of 3.26% drop rate can be obtained. Keywords: Taguchi method; parameter optimization; performance evaluation; ad hoc network; dynamic source routing protocol.