Power Control Mechanism Research Based On Game Theory For Cognitive Radio Network

The development of wireless communication systems has led to an increasing demand for spectrum resources,and how to improve the utilization of spectrum resources is becoming more and more important.Cognitive radio networks can provide secondary users(SUs)with access to primary users(PUs)spectrum channels and significantly improve the spectrum efficiency of the network system by sharing spectrum in the underlying system.In order to ensure the maximum performance of the system,it is necessary to design an effective spectrum-sharing scheme,so that the SUs has the opportunity to use the idle spectrum without affecting the normal communication of the PU.Power control is an effective resource allocation strategy,which can limit the transmission power of the SUs and ensure that the PUs can communicate normally under the interference power constraints(IPC).In this thesis,we utilize game theory to analyze and study the power control problem,and do some researches deeply in two aspects shown as follows:1.To deal with the problems of interference existing among users and powerconsumption caused by the huge transmission power requirement of the SUs in Cognitive Radio Network,a novel power control algorithm based on Stackelberg game theory is proposed,which helps to control the power.A two-layer network model with one PU and some SUs coexisting is built firstly,in this network model,the PU acts as leader,the SUs act as followers,then a quantitative analysis of total interference generated by the SUs is made.In the process of multiple games played among the PU and SUs,the PU adjusts the price of unit interference dynamically,under the premise of normal communication,to ensure maximizing revenue and increasing the enthusiasm for participation as much as possible.The simulation results show that,with a stable service quality provided by the PU and SUs,the proposed algorithm not only decreases the transmistting power of the SUs effectively but also improves the system performance.2.When multiple-user resources allocated in cognitive radio networks,a large amount of channels and power strategy information are needed to interact,which will cause a large occupation and expend of system resouses.To solve this problem,we analyze the users with a non-cooperative game model and propose a non-cooperative game model based on Q-learning algorithm to realize the joint channel selection and power control dynamically.In the process of self-learning,the users will observe their own rewards and do Q-learning with an unified strategy,the learning result gradually converges to the optimal strategy of optimal channel and power allocation.As simulation results shown that the algorithm can converge to Nash equilibrium with high probability,and the users can get higher system capacity after channel selection.