Research On Energy Efficiency Based Beamforming Technique In Cloud Radio Access Network

With the development of the wireless communication technologies,the fifth generation mobile communication(5G)should provide not only higher data transmission rate and better quality of service for mobile users,but also reduces system overhead to realize green communications.However,to achieve higher spectrum efficiency and provide higher data rate for mobile users,the radio resources should be allocated and optimized more efficiently.Meanwhile,in the current network scenarios,the ratio of the base station(BS)power consumption and the overall power consumption accounts for more than 50%even adopting micro-cell or femto-cell BSs.Thus,to achieve green communications,it is necessary to measure new metric for green communication,and deploy more reasonable network topology,and propose low power network structure,as well as to adopt some new technologies to improve resource efficiency.As a new wireless access network architecture,cloud radio access network(C-RAN)utilizes cloud computing,and adopts advanced technologies such as centralization,cooperation,cloud,green and so on,and is regarded as one key standard to achieve green communications.To utilize the resource more efficiently,C-RAN usually processes the resources centrally according to the requirements of users as well as the network status.Coordinated multi-point(Co MP)transmission and reception technique can transform the interference signals into useful signals in order to suppress the co-channel interference.Therefore,by introducing Co MP in C-RAN and designing cooperative beamforming,the received signal interference noise ratio(SINR)can be largely improved and thus the transmission performance is boosted.Nevertheless,except for improving the system resource efficiency,C-RAN brings a series of challenges for wireless networks as well,and the most direct one is energy consumption.Towards this end,the key goal of C-RAN is to study the tradeoff between spectrum efficiency and power consumption,and achieve the promotion of energy efficiency(EE).In the practical communication system,due to limited bandwidth feedback,doppler shift,inaccuracy of channel estimation and pilot pollution,the obtained channel state information(CSI)is usually imperfect,and the imperfection in CSI deteriorates the system performance.Therefore,it is necessary to design robust beamforming due to the imperfection of obtained CSI.Based on the above concerns,this dissertation focuses on the robust beamforming design under imperfect CSI to maximize the energy efficiency in downlink C-RANs.Firstly,in the case of fixing the users' rates,this dissertation focuses on beamforming design problem for power minimization under imperfect CSI in C-RAN,and this dissertation solves this problem through jointly optimizing BS mode,user association and robust beamforming.By adopting the group sparse optimization theory in compress sensing,this dissertation proposes an iterative group sparse bemforming(IGSBF-SDP)algorithm to solve the optimization problem via solving a series of semidifinite programming(SDP)problems,and the convergence of this IGSBF-SDP algorithm is also verified.To reduce the computational complexity,this dissertation proposes to transform the problem into a second order cone programming(SOCP)one and proposes an iterative difference of convex function(IDC)based SOCP heuristic two-step selection(IDC-SOCP)algorithm.This IDC-SOCP algorithm combines the IDC and bisection search to determine BS mode and user association separately.At last,this dissertation verifies the performance of the proposed algorithms through the aspects of convergence rate,power consumption and impact of channel errors on the total power consumption.Secondly,this dissertation focuses on the beamforming design problem for maximizing the energy efficiency under imperfect CSI and constrained fronthaul capacity in C-RAN,and this dissertation tries to solve it via jointly optimizing beamforming and fronthaul compression.To obtain the optimal solution of the considered problem,this dissertation first proposes a branch-and-bound based algorithm.Although this optimal algorithm has very high computational complexity,it provides the upper bound performance to any other suboptimal algorithms.To reduce the computational complexity,this dissertation proposes a Dinkelbach based two-layer distribution(TLD)iterative algorithm.Specifically,the outer layer updates the energy efficiency through a bisection search procedure,while the inner layer combines the block coordinate decent(BCD)and alternating direction method of multipliers(ADMM)algorithms,and returns the closed-form solution in a parallel manner.At last,this dissertation analyzes the maximum BS transmit power,fronthaul capacity and channel error on the energy efficiency to verify the effectiveness of the proposed algorithms.Finally,for the C-RAN assisted secure communication systems,due to the broadcast properties of wireless channels,the eavesdropping receivers(ERs)can receive and decode the information transmitted by the BSs and they may also interfere information receivers(IRs)via transmitting signals.Thus,this dissertation introduces artificial noise(AN)to interfere ERs in order to improve the network security performance from the perspective of information theory.This dissertation investigates the energy efficient beamforming design problem under imperfect CSI in a secure C-RAN system,and this dissertation jointly optimizes beamforming and AN to solve it.To begin with,this dissertation proposes a branch-and-bound beamforming algorithm to obtain the optimal solution of the considered optimization problem.Due to the high computational complexity of the optimal solution,this dissertation proposes a sequential convex approximation(SCA)based iterative beamforming algorithm.The SCA algorithm first adopts the epigraphical transformation of the original problem,and then linearizes the non-convex constraints by the first-order Talyor expansion.The imperfect CSI related constraints are reformulated into convex ones by using the S lemma,and the original problem is then solved by an iteration procedure.Simulation results show that the energy efficiency of the proposed algorithm approximates the optimal one while guaranteing secure communications.