Spectral And Energy Efficiency Optimization Of Cell-free Massive MIMO Systems

With the rapid development of mobile communication systems,spectral efficiency(SE)and energy efficiency(EE)are the key metric for the design of the fifth-generation mobile communication system(5G).Due to the limited spectrum resources and the demand for high-rate capacity,as a traditional performance index,SE has already received extensive research.Considering improving the SE and EE,the cell-free Massive Multiple Iutput Multiple Output(MIMO)system as an important branch of the distributed massive antenna systems,which exerts its unique advantages and has the potential to be a key technology for mobile communications.Based on the user-centered mode and adopted time division duplex(TDD)protocol,its aims to make a large number of distributed wireless access points(APs)cooperate with each other in the network,and coherently serve all users simultaneously with uniformly good quality-of-service(Qo S)in same time-frequency resource.Meanwhile,all APs are connected to the central processing unit(CPU)which responsible for the coordination and data processing via an error-free limited-capacity of backhaul links network.Compared with cellular massive MIMO system,the cell-free Massive MIMO system eliminates the concept of cell boundaries,decreasing the distance between the APs and the users,avoiding handover frequently and reducing the path loss,which suppressing multiuser interference effectively,having higher coverage and huge macro diversity.However,the scalability and performance of the system are limited significantly by the backhaul capacity and interference.Based on the latest research status of mobile communication development at home and abroad,this thesis investigates the power allocation optimization of SE and EE of the cell-free Massive MIMO system.First of all,full-pilot Zero-Forcing(fp ZF)precoding algorithm as fully distributed signal processing scheme,where APs uses the locally obtained channel state information(CSI)to operate detection/precoding locally.The algorithm provides interference cancellation gain and enhances the expected signal without additional backhaul overhead,which actively suppressing the multiuser interference without sharing instantaneous CSI between the APs and the CPU and among APs.Secondly,considering the Rayleigh fading channel model,the influence of channel estimation error,power control and pilot contamination,it derives the achievable closed-form expression of SE for uplink and downlink.Taking each user Qo S and power constraint into account,the power control optimization problem of SE is modeled,and SE is maximized by selecting the best power control coefficient.Finally,the achievable closed-form EE expression is derived considering the power consumption,uplink and downlink EE power allocation optimization problems is modeled.We propose an optimal power control algorithm by finding the optimal power control coefficients under each user Qo S constraint and power constraint to maximize total EE.Furthermore,two simple AP selection schemes is proposed: received-power-based selection and largest-large-scale-fading-based,which make system convenient for implementation,can be implemented through fp ZF scalability,and there is no limit to the number of users that can be served,providing reasonable and good Qo S for all users.And the simulation results indicate that the overall EE of the system is increased.Importantly,we proposed a novel Sequential Convex Approximation(SCA)algorithm to solve the non-convex power control optimization problem.SCA strategy is a continuous geometric programming algorithm,which can be determined by transforming alternating convex programming.Simulation results demonstrate that the proposed SCA algorithm can converge to a stable value after a few iterations and outperform the equal/full power transmission counterpart.The proposed algorithm has superiority in improving the total SE and total EE.