Game Theory Based Network Selection In Heterogeneous Wireless Networks

With the development of mobile communication technologies, multiple wireless access technology will coexist and gradually merge in the future. Different wireless access technologies exhibit different features on bandwidth, delay, coverage and mobility, and meet different requirements of different users. Access network selection is a key technology, which provides the best service for users, ensure network balance and makes full use of radio resources of all networks. Therefore, this thesis studies the network selection based on game theory.In multi-service, multi-user and multi-network scenario, we propose an adaptive network selection algorithm based on evolutionary game. Users divided into different collection based on optional network sets and service types, and dynamically adjust network selection strategy. The algorithm considers the difference of networks and services. Dynamic algorithm prevents the occurrence of network congestion. The simulation results show that the proposed algorithm can make network selection strategy be stable as selecting the appropriate adjustment rate.In scenario that network access networks are independent of each other, we propose a heterogeneous wireless access network selection algorithm based on Stackelberg game. In this algorithm, transmission rate, bandwidth, delay and the master-slave feature of network selection are considered. This algorithm using linear cost function and utility function to ensure convergence. Simulation results show that this algorithm can rapidly make the user-network transmission rate strategy to achieve Nash equilibrium, and increase net utility function value largely.