Game Theoretic Approach for Network Access Control in Heterogeneous Networks

The latest mobile communication devices are usually equipped with multiple network interfaces with the capability of accessing multiple types of wireless networks. Since each network has limited capacity, a network may experience congestion if too many mobile devices access this network. To avoid congestion, an intelligent distributed network selection scheme is needed for each mobile device. In this paper, we propose a game theoretic approach to the problem of network selection. Specifically, we formulate the problem as a repeated stochastic game with a goal of maximizing total utility while achieving high proportional fairness among all users in the service area. To solve this problem, we develop a Lyapunov optimization algorithm to obtain the optimal action for each user from the game manager perspective. Simulation results show that our proposed approach can achieve much higher utility than existing schemes while maintaining good fairness index.