DSR protocol scalability optimization in mobile ad-hoc networks with ANFIS

Abstract

An ad-hoc network is a collection of mobile nodes, which are forming a network without any infrastructure established. To help these kind of networks there exist a variety of protocols that give us an administration of these networks. In our case we are working over the DSR protocol (dynamic source routing). We work with the route error optimization problem in ad hoc hierarchical networks. An ad-hoc wireless network is a collection of mobile wireless nodes that form a temporal network without any established infrastructure or centralized administration. The principal problem to resolve is about how many nodes the DSR protocol can accept without putting down their behavior. We observed that if the nodes are as bigger as the network size their behavior is going impair, so we are looking for the ad hoc network that is capable of working the same way no matter how many nodes are there in the network. To resolve the problem we propose to use an ANFIS (adaptive neuro-fuzzy inference system), which is working inside the network. The protocol that we use is the DSR, which is used for ad-hoc networks, based on the concept of source route. The results that we have obtained were implemented in the NS (network simulator, discrete event simulator to networks research).