Research On Relay Selection And Resource Allocation Schemes In Cooperative Relay Networks

Cooperation and relay are one of the key technologies of the future wireless networks, which draw wide attention from both academic and industry field. Due to the existence of relay nodes, the resource allocation and relay selection schemes in cooperative relay networks becomes more complicated. With the growing demand of wireless network capacity and the concern of wireless network energy, spectral efficiency and energy efficiency become a hot topic. In cooperative relay networks, how to complete resource allocation and relay selection effectively, achieving energy efficiency and network capacity optimization, guaranteeing fairness among users, is worthy of further study.The work was supported by Special Issue of Major National Science and Technology,’TD-LTE-Advanced Wireless Relay Test Equipment Development’, and Outstanding Graduate Innovation Fund of the School of Information and Communication Engineering, Beijing University of Posts and Telecommunications’The Research of Energy Efficient Transmission Technology in Wireless Relay Networks’.This thesis focuses on relay technology, and makes research on resource allocation scheme in relay enhanced cellular networks and relay selection scheme in relay Ad hoc networks, designes new resource allocation and relay selection schemes.Firstly, this thesis reviews the resource allocation and relay selection techniques in cooperative relay networks, and makes summary and analysis on existing research. In relay enhanced cellular networks, graph coloring based resource allocation scheme is proposed. Simulation results show that the proposed scheme can improve network capacity and guarantee fairness among users. In relay Ad hoc networks, artificial intelligence based relay selection schemes are proposed. In single-user (single source and single destination) scenarios, both single-objective optimization and multi-objective optimization models are proposed. Discrete quantum particle swarm optimization and the concept of non-dominated sorting are used to solve the problems. Simulation results show that the proposed algorithm can improve network capacity and energy efficiency. In multi-user (multiple sources and multiple destinations) scenarios, both single-objective optimization and multi-objective optimization models are proposed. Continuous quantum particle swarm optimization and the concept of non-dominated sorting are used to solve the problems. Simulation results show that the proposed schemes can improve Max-Average-Reward, Max-Proportional-Fair and Max-Min-Reward.