Research On Game Theory Based Spectrum Sharing In Cognitive Radio Networks

In recent years,with the rapid development of mobile communication,the demand for wireless spectrum is growing fast.The contradiction between spectrum scarcity and low efficiency becomes increasingly important under the existing spectrum authorization mechanism.Cognitive radio,as a wireless technology for effective use of spectrum resource,can reduce the constraints of spectrum and bandwidth and promote the development of wireless technology.Cognitive radio can achieve multi-dimensional spectrum reuse by dynamic detecting the idle spectrum and then solve the contradiction between spectrum scarcity and low efficiency.The users in cognitive radio network have the ability of observation,learning and making optimal decisions,thus the behavior model of unconditional cooperation is not suitable.As a useful tool for analyzing the interaction of wireless users in cognitive radio network,game theory has dominant advantages in modeling and analyzing.This dissertation mainly focuses on spectrum sharing in cognitive radio network.The main contribution of our work lies in proposing novel solutions and designing schemes for resource allocation of different application scenarios in cognitive radio network using game theory and its branches,optimization theory and probability theory based on game strategy analysis of cognitive users.The performance of the proposed algorithm and mechanism is analyzed through the computer simulation and provides a powerful theoretical support for flexible and efficient spectrum sharing.The main innovations are as follows:Firstly,the relationship between spectrum utilization and QoS of the primary user has been studied with the imperfect sensing information and also the influence of the price mechanism on the secondary users' strategy.A distributed joint spectrum detection and power allocation algorithm is designed for maximizing the downlink throughput of secondary networks.The proposed algorithm converges fast to a NE by parallel update and the convergence is proved by constructing Nikaido-Isoda function.The price mechanism reduces the interference among secondary base stations and improves the system capacity.Secondly,the problem of energy limited of cognitive radio network has been investigated.Under the constraint of guaranteeing the QoS of the primary user the uplink energy-efficient resource allocation in cognitive radio celluar networks has been solved by exploiting the properties of the nonlinear fractional programming.A double iteration scheme has been employed to maximize the energy efficiency.Simulation results show that the convergence of the proposed algorithm is quick and the scheme brings higher EE than that of the spectrum efficiency optimization algorithm and the random allocation algorithm.Thirdly,the mechanism design of multiple relay selection and power allocation has been studied.And Stackelberg game model is employed to consider the benefits of the source node and the relay nodes to maximize their utilities.A gradient based iteration algorithm is proposed to obtain the Stackelberg equilibrium solution to the distributed resource allocation problem.Numerical simulation results show that the convergence of the mechanism is insensitive to the number of relay nodes and can maintain good performance in large-scale distributed scenarios.Finally,quick and efficient mechanism design of dynamic spectrum leasing has been investigated and the requirements of budget constraints and resource integrity make it more applicable in the practical system.Based on the multi-object auction a modified uniform price mechanism with true-telling property is proposed and can allocate the channel resources quickly and effectively.The incentive compatibility design improves the robustness of the system against malicious bid nodes.Low computational complexity and signaling interaction provide a feasible way for the distributed realization of cognitive radio networks.