Research On Energy Efficiency Optimization For Massive MIMO Systems In 5G

Nowadays,with the rapid increase in demand for massive intelligent devices and high-quality user experience,the wireless broadband traffic and wireless transmission rate have shown a trend of exponential growth.However,the 4th generation mobile communication system that currently support transmission rates up to 1 Gbps(Gigabits per second)is obviously unable to meet future communications needs.Therefore,how to improve and optimize the spectrum efficiency of the 5th generation mobile communication system(5G)has become the latest research trend,the academia,industry,and various communication organizations are trying to grab the initiative in the future communications field.Meanwhile,as the number of user increases and the size of the network expands,the energy consumption of the wireless communication network also increases sharply.The wireless communication system of greenhouse gas emissions caused by environmental pollution and huge energy consumption have been associated with caused extensive concern of the operators and the society as a whole.So,a green communication system,which goal is to maximize energy efficiency,has become a development direction of the future wireless mobile communication.Among the existing key technologies of 5G,massive multiple-input and multiple-output(MIMO)technology is regarded as one of the most potential candidate technologies for 5G due to its excellent propagation channel and high spectrum efficiency.The core idea of massive MIMO technology is to obtain spatial diversity and spatial multiplexing gain by deploying antenna array at the base station,so as to greatly improve the spectrum efficiency of the system.In addition,applying massive MIMO technology to heterogeneous networks can further meet the high performance requirements of 5G networks.Based on the above background,the research on energy efficient optimization technology of massive MIMO systems in 5G is a key problem that needs to be solved urgently and is of great significance.The main research contents of this thesis are as follows.First,we analyze the performance of a massive MIMO downlink system.Under the assumption that both the base station and the user have the perfect channel state information(CSI),a single-cell massive MIMO downlink system model is established based on a typical multi-user MIMO system.By using maximum-ratio transmission(MRT)precoding,we respectively derive three tractable closed-form expressions that corresponding to the lower bound,approximation and upper bound on the achievable rate.Based on the proposed closed-form expressions,the power efficiency of the system is investigated as the number of transmit antennas increases.Second,we study the energy efficient optimization of a massive MIMO downlink system.Assuming that both the base station and the user have the perfect CSI,a joint antenna selection and power allocation scheme is proposed to maximize the system energy efficiency.Based on a tractable lower bound of the achievable downlink rate with linear zero-forcing(ZF)precoding,a non-convex energy efficiency optimization problem is formulated with non-linear constraints.Since the optimization variables are highly interrelated,it is impractical to directly derive the closed-form solution to the original problem.To solve this problem,an effective iterative algorithm that maximizes energy efficiency is proposed according to the Lagrangian dual analysis,where the optimal number of transmit antennas and power allocation are solved alternately until convergence.Third,we propose a power allocation scheme for a massive MIMO downlink system with imperfect CSI.Assuming that the base station has imperfect CSI,while the user has perfect CSI,a channel estimation model is constructed based on the prior knowledge of estimation error.Considering two linear precoding strategies at base station,i.e.,MRT and ZF,we first derive the corresponding approximate closed-form of the achievable sum-rate under imperfect channel state information.Furthermore,an energy-efficient power allocation scheme is formulated as a non-convex optimization problem.An iterative power allocation algorithm that maximizes the system energy efficiency is developed based on the Lagrangian dual approach.Next,by assuming that both the base station and the user have imperfect CSI simultaneously,we first show a lower bound on the ergodic achievable downlink rate,which can be applicable for any precoding scheme.Using MRT and ZF precoding,the tractable closed-form expressions are derived for the proposed lower bounds.With these closed-forms,we formulate the energy-efficient optimization problems under minimum transmission rate and maximum transmit power constraints.Since the proposed optimization problem with MRT is non-convex,we transform the original non-convex objective function into two sub-functions and prove the convexity of two sub-functions.By employing the difference of convex programming,we develop an energy-efficient power allocation algorithm that maximizes the energy efficiency.When the proposed algorithm converges,the sub-optimal power allocation solution to the non-convex optimization problem is obtained.Finally,we investigate the ?-fairness network utility optimization problem of joint user association and power allocation for a downlink Massive MIMO heterogeneous network with imperfect CSI.By utilizing ZF beamforming,we show a new closed-form lower bound expression on the ergodic achievable rate.Furthermore,we formulate the optimization problem as a mixed-integer nonlinear programming problem,which is non-convex and NP-hard,to achieve the maximization of a-fairness network utility.Consequently,a joint iterative algorithm with respect to user association and power allocation is developed by decomposing the original problem and relaxing the constrains.