Application of game theory in wireless communication networks

The ability to model independent decision makers whose actions potentially affect other decision makers makes game theory attractive to analyze the performances of wireless communication systems. Recently, there has been growing interest in adopting game theoretic methods to wireless networks for power control, rate adaptation and channel access schemes. This thesis focuses the application of dynamic game theory and mechanism design in cognitive radio networks, Wireless Local Area Networks (WLAN) and Long Term Evolution (LTE) systems. The first part of the thesis aims to optimize the system performance through the transmission rate adaptation among wireless network users. The optimal transmission policy of each user is analyzed by formulating such a problem under general-sum Markov game framework. Structural results are obtained and provably convergent stochastic approximation algorithm that can estimate the optimal transmission policies are proposed. Especially, in the switching control Markov game theoretic rate adaptation formulation, it is proved that the optimal transmission policies are monotone in channel state and there exists a Nash equilibrium at which every user deploys a monotone transmission policy. This structural result leads to a particularly efficient stochastic-approximation-based adaptive learning algorithm and a simple distributed implementation of the transmission rate control. This thesis also considers the spectrum allocation problem in an LTE system. By incorporating cognitive capacities into femtocell base stations, the Home Node-Bs (HNBs) can be formulated as cognitive base stations competing for the spectrum resource while trying to minimize the interference introduced to the evolved Node-B (eNB) which is also referred as primary base station. Given the primary base station spectrum occupancy, the spectrum allocation problem among HNBs can be formulated under game theoretic framework. The correlated equilibrium solution of such a problem is being investigated. A distributed Resource Block (RB) access algorithm and a correlated equilibrium Q-learning algorithm are proposed to compute the spectrum allocation solutions under static and dynamic environments, respectively. The last part of the thesis uses mechanism design to design a truth revealing opportunistic scheduling system in a cognitive radio system. A mechanism learning algorithm is provided for users to learn the mechanism and to obtain the Nash equilibrium policy.Applied Science, Faculty ofElectrical and Computer Engineering, Department ofGraduat