Abstract
Modern Wireless Local Area Networks (WLANs) with relatively high data rates have become an attractive technology for providing Internet connectivity for mobile users. Ad-hoc networks are a collection of mobile nodes that can be deployed without the need for any centralized management infrastructure. In such a set-up, to establish the required communication paths, each node must be willing to act as a potential router. In practice though, some nodes may act selfishly and refuse to forward packets. In Ad-hoc networks, a node may be considered as misbehaving for different reasons, for instance when it acts selfishly, refusing to forward packets. In some circumstances, the node can be overloaded, or they simply want to save their resources by not forwarding packets unless they are of direct interest to the node itself. Conversely, these nodes may still be expecting others to forward packets on their behalf In this paper, we report the experimental results obtained from a typical Ad-hoc networks that contain selfish nodes. We also analyze the behavior of the nodes, to establish some quantifiable measure of their reliability. Such measures, based on the behavior history of the nodes, are then utilized to improve the performance and reliability of the widely used Ad-hoc On Demand Distance Vector routing protocol. We also report the results of simulations of large Ad-hoc networks in the presence of malicious or selfish nodes. These results clearly indicate the capabilities of the proposed approach in discovering communication paths with a minimal number of malicious or selfish nodes.