Robust Power Control In Cognitive Radio Networks With Uncertain Channel Gains

With the fast development of information society, the demand for mobilecommunication and broadband wireless access are growing. So, wireless spectrumresource becomes very valuable. Cognitive radio (CR) has been proposed as an excitingemerging technology in this years. It has the potential to solve the problem of scarcespectrum resource. In order to improve the low frequency spectrum utilization, CR sharethe spectrum with primary user. But the interference to the primary users is avoided orbeyond limited to an acceptable level. Therefore, power control plays a key role inreducing interference in wireless communications.In cognitive radio networks, most studies of power control algorithm are under theassumption that the parameters of the objective function and constraints are constant orperfectly known. However, in practical systems, the parameters are imperfectly known ortime varying. If these algorithms deal with the uncertainty of parameters，they may lead toperformance degradation in practical systems. So, robust optimization has been a quitenew research topic in dealing with uncertain parameters.Firstly, for the channel gains uncertainty of power control problem in cognitive radiosnetworks. we design a robust power control algorithm under the worse-case. With theuncertainty of channel gain, in order to satisfy the cognitive users and primary users’ Qos,the improved DCPC and GDCPC approachs are applied to power allocation. Simulationresult shows that the proposed method improves the cognitive users’ over performancethan tradition method with the interference limited to primary user.Secondly, for maximizing cognitive users’ utility, we establish a chance-contraintrobust power control optimization problem in cognitive radios, under the cognitive usersand primary users’ outage probability limited. A gradient algorithm is used to obtainoptimization power, and the convergence of the gradient algorithm is proved. Simulationresults show that, the algorithm not only can overcome the uncertainty but also canimproves the cognitive users’ performance.Finally, we consider the high load in cognitive networks. In order to increase the number of the cognitive users assess to the networks, we combine the standardinterference function and maximum power adaptation approach to obtain optimizationpower of cognitive users. Moreover, admission control is introduced to impove the accessusers’ Qos and system reasources can be more reasonable utilization. Simulation resultshows that the schemes is convergence and improve the average SINR.