Measuring uncertainty of MANET

Abstract

MANET is a multi-hop network comprised of nodes that need to work in collaboration with each other for effective routing. But unfortunately, there is possibility of some nodes misbehaving for their own interests and disturbing the network. Therefore, reliability among nodes is very important in the network. Uncertainty is of prior importance, which deeply impacts a node's anticipation of others behavior and decision making during interaction. In this paper, we introduce the concept of uncertainty to rationally evaluate trust. We analyze the impact of network size, malicious nodes, number of connections and speed on the uncertainty of network under two cases i.e. with and without recommendations. The paper contributes to find uncertainty index of every node and ultimately obtaining average uncertainty of network. This is achieved by mathematical theory of evidence. Thus we conclude, uncertainty with recommendation is less as compared to uncertainty without recommendation.