Detection and mitigation of misbehavior in cooperative wireless communications

Cooperative diversity protocols are designed with the assumption that wireless users always cooperate in a socially efficient manner. This assumption may not be valid in commercial wireless networks where users may violate rules of cooperation for malicious or selfish reasons. In the presence of misbehavior, it is difficult to ensure a socially efficient cooperation without techniques to enforce cooperation. In this dissertation, we develop techniques to detect and mitigate misbehavior in cooperative wireless networks. We first examine effects of misbehavior on performance of existing cooperative protocols. We show both analytically and by simulation the incurred performance degradation. We show based on game theoretic arguments that existing cooperative schemes are characterized by a non-cooperative Nash equilibrium. We examine using evolutionary game dynamics effects of a small number of mutants on the behavior of a population of cooperators. We then propose two classes of misbehavior detection techniques. The first class of detectors determine behavior of relay nodes using signals transmitted over source-destination channel as a side information. The second class of detectors are developed for cooperative networks with hybrid automatic repeat request (HARQ) protocol. We develop optimal detection techniques (1) based on uniformly most powerful (UMP) test; (2) based on sequential probability ratio test (SPRT). Misbehavior is determined based on control packets exchanged to reveal status of transmissions. Performance of the proposed detection techniques is evaluated based on detection delay, computational complexity and memory requirement. The proposed detection techniques provide only a localized view of the cooperative system. However, to mitigate further effects of misbehavior a mechanism is required to give each node a global view of the network. To this end, we formulate cooperative communications as a dynamic game with incomplete information, which provides a framework (1) to develop a model for selfish behavior and (2) for designing a reliable partner selection scheme in the presence of uncertainty. We show that the proposed dynamic Bayesian game model captures vital aspects of cooperative communications. We show that the proposed dynamic Bayesian game satisfies the requirements for the existence of Perfect Bayesian Equilibrium.