Joint Beamforming Design Of MIMO System Based On Intelligent Reflecting Surface

As the fifth-generation wireless networks are deployed worldwide,traditional communication technologies need help meeting the requirements of increasingly higher data rates,energy efficiency,and spectrum utilization.Intelligent Reflecting Surface(IRS)technology has been proposed to address the above challenges.By designing corresponding beamforming techniques,IRS can flexibly regulate electromagnetic waves to improve wireless communication’s cost efficiency and system performance.In this thesis,IRS is applied to the traditional Multiple Input Multiple Output(MIMO)communication systems,and the optimization of the system’s channel capacity is studied.Firstly,for the new wireless communication technology paradigm of the IRS,this thesis focuses on its fundamental principles,introducing its structure and control principles,typical architecture,and application scenarios.After elucidating its ability to build an intelligent radio environment,the capacity problem of IRS-assisted MIMO communication systems is modeled.Relevant optimization theories are introduced,providing theoretical support for subsequent joint beamforming design work.Secondly,the problem of maximizing the channel capacity of an IRS-assisted massive MIMO communication system with antenna selection is considered.This thesis proposes an efficient alternating optimization framework for non-convex constraints and coupled optimization variables,which decomposes the system capacity optimization problem into antenna selection and joint beamforming subproblems.A low-complexity antenna selection algorithm is proposed by approximating the objective function using the Sum-Path-Gain Maximization criterion.Based on the unique structure of the objective function,a passive beamforming algorithm based on Block Coordinate Descent is further proposed,iteratively solving the optimal phase of the IRS reflecting element.The simulation results show that the alternating optimization framework not only reduces the hardware cost of the base station but also improves system performance and reduces computational complexity of each iteration of passive beamforming to O(N(N-1))with acceptable performance loss.Finally,in order to reduce the hardware cost and implementation complexity of the IRS reflecting elements,this thesis designs a discrete phase joint beamforming scheme to optimize the capacity of the IRSassisted MIMO communication system.An alternating optimization algorithm is used to iteratively optimize the transmission beamforming and the phase of each reflecting element,and a nearest point projection algorithm is proposed to project the optimal continuous phase of the reflecting element onto a discrete phase set.Based on the time scale,an approximation algorithm is proposed,which reduces the search space of the optimal discrete phase and reduces the complexity of passive beamforming to O(N2).The simulation results show that the passive beamforming scheme can achieve channel capacity close to continuous IRS with a lower discrete quantization level and achieve the balance between hardware cost and system performance,which provides a valuable reference for the engineering practice of IRS.