Enhancing Benign User Cooperation in the Presence of Malicious Adversaries in Ad Hoc Networks

Decentralized and unstructured networks are becoming more prevalent today (e.g. ad hoc networks). Like every network, they depend on the cooperation of their users to survive. However, each user does not necessarily know who the others are, or what their intentions are. Since there is no centralized infrastructure, the users can only base their decision on what they observe themselves. Ideally, they would like to cooperate only with users that have common interests with them. In this paper, we use a game theoretic model for the above situation. We assume there are only two kinds of users, good (benign) and bad (malicious). Good users receive a high game theoretic payoff when they cooperate with other good users, but a low payoff when they cooperate with bad users. We propose behavior rules (strategies) to achieve equilibria that enable as many good users as possible to cooperate with each other, and at the same time minimize the number of good-bad cooperations