Research On Power Allocation Of Jointing Energy Efficiency And Spectrum Efficiency Optimization In Distributed Cognitive Radio Networks

Recently, with the wideband service requirements of global mobile users dramatically increasing, the global mobile data traffics explosive grow. Meanwhile, the next generation wireless communications is also facing daunting challenges: scarcity of limited spectrum resources, the rapid growth in wideband demanding and energy consumption. Therefore, in order to efficiently use the limited spectrum resources to meet people’s wonderful life in wideband world and achieve a sustainable future of wireless communication, cognitive radio technology come into people’s eyes. In cognitive radio networks, secondary users(SUs) can intelligently sense the cognitive wireless environment and automatically adjust its parameters to send information without changing their hardware conditions. In the problem of wireless resource allocation over the multi-domain: "user", "time", "frequency" and "airspace", how to improve the system spectrum efficiency and prolong the life of equipment has become a hot topic. In this thesis, power control is employed for SUs over "frequency domain", "airspace" and "user domain" and study the problem of energy efficiency(EE) and spectral efficiency(SE) in cognitive radio(CR) networks. The main works are as following:(1) The problem of joint power and rate optimization in multi-channel heterogeneous cognitive radio networks(H-CRNs) is studied, and the Leader-Follower relationship between primary users(PUs) in macrocells and SUs in femtocells is modeled as a Stackelberg game. Since the coupled relation between PUs and SUs, PUs set price for each sub-channel and adjust the price of sub-channels to control the interference from SUs below an interference threshold. In Stackelberg game, power allocation based on non-cooperative game theory is applied to each PU to maximize their rates, while for each SU, jointing power and rate optimization based on market competition is taken into account to maximize their individual utility. Furthermore, a Primary systems and Cognitive systems Joint Resource Allocation algorithm(PCJRA) is proposed to maximize the utilities of SUs subject to the maximum rate requirements for PUs. Meanwhile, the existence and uniqueness of Nash Equilibrium(NE) in PUs are proved, and the conditions for the uniqueness of Competitive Equilibrium(CE) and the convergence of the distributed algorithm in SUs have also been derived. Theoretical analysis and simulation results demonstrate both the feasibility of the multi-channel H-CRNs model and the effectiveness of the proposed algorithm.(2) The EE optimization problem for multi-user cognitive radio over MIMO interference channel(IC) is modeled as a non-cooperative game, which aims to maximize the EE of the multi-antenna SUs in terms of the number of bits transmitted per unit energy consumption(joule). First, the corresponding EE optimization problem which is a non-convex multi-objective fractional problem is transformed into a multi-objective non-fractional problem. Second, the primal decomposition is taken into consideration to relax the temperature-interference threshold. Then, by introducing the Lagrange factor for the CR constraints and based on Lagrange method, the multi-objective with multi-constraints optimization problem is converted to multi-objective without constraints optimization problem. Last, it is simplified and power values are derived by exploring the matrix decomposition and pre-coding theory. Furthermore, a distributed iterative EE optimization algorithm(DIEEOA) for the multi-user cognitive radio over MIMO IC is proposed to obtain the Nash Equilibrium(NE) of the non-cooperative game. Effectiveness of the algorithm is validated through matlab simulation, and the system parameters’ impact on the EE is discussed.(3) For the multiusers CR MIMO IC systems under imperfect channel state information(CSI), considering the channel estimate error the EE optimization of multiuser under the constraints of the interference temperature threshold and the maximum transmission power is constructed as a non-cooperation game. To solve the problem, the scheme of Energy efficiency optimization combined with precoding for multiusers CR MIMO IC under imperfect channel state information is formulated, which maximizes the EE of each SUs in the multiusers CR MIMO IC systems. Simulation results show that the precoding can reduce the effect of imperfect CSI on the energy efficiency of systems under the optimization of energy efficiency, furthermore, which can significantly improve the energy efficiency of the systems under imperfect CSI.