Research On Intelligent Reflecting Surface Assisted Robust Joint Transmission Beamforming Design

One of the fundamental challenges in achieving ultra-large capacity and ultra-reliable wireless communication is the random,time-varying and uncontrollable wireless communication environment.Some efforts have been devoted to developing various wireless communication technologies to compensate for wireless channel fading or adapt to dynamic channel conditions.However,the traditional optimization methods mainly focus on the design of the transceiver,and the control of radio environment is limited.Recently,Intelligent Reflecting Surface(IRS)has been recognized as a promising technology for reconfiguring wireless propagation environments by controlling signal reflection.Each element of IRS can independently induce amplitude and phase changes of the incident signal,thus synergistically realizing reflected beamforming for directional signal enhancement or destructive attenuation,which is expected to revolutionize the current network optimization model.By correctly designing the IRS reflection coefficient,it has been proved that IRS can significantly improve the wireless applications’ performance.However,the performance gain of an IRS-assisted communication system depends on the accuracy of Channel State Information(CSI)of IRS-related links.In practical applications,due to the passive and large number of elements of IRS,channel estimation errors are inevitable.Therefore,to realize the potential of IRS-assisted communication systems,it is necessary to study robust optimization of Base Station(BS)transmit beamforming and IRS reflection coefficient for the imperfect estimation of CSI.Therefore,this thesis investigates the IRS-assisted robust joint transmission beamforming design scheme,and the main contents include:1.Aiming at the channel estimation error problem of BS-IRS-user cascaded channel estimation methods,this thesis proposes a robust joint beamforming scheme for an IRS-assisted Joint Processing Coordinated Multipoint multi-cell multi-user system,which can guarantee system performance in the worst case of CSI bounded error.Specifically,user data is available at all BSs,which can jointly provide services for cell-edge users.To save energy consumption,the objective is to minimize the total transmit power by jointly optimizing the transmit beamforming at the BSs and the phase shift at the IRS,while satisfying the BSs’ individual power constraints and users’ worst case target rate constraint.To solve this complex non-convex problem,this thesis proposes an alternating optimization(AO)algorithm based on semidefinite relaxation(SDR)and convex approximate transformation.Simulation results show that the proposed robust joint beamforming scheme can converge rapidly and outperforms the reference scheme.2.Aiming at the error problem of the channel estimation methods based on separate estimation of IRS-related links,this thesis proposes a robust coordinated beamforming scheme for an IRS-assisted Coordinated Scheduling/Coordinated Beamforming Coordinated Multipoint multi-cell multi-user system.Specifically,to save fronthaul capacity,CSI of cell-edge users is shared among BSs,while users’ data is only available at the BS that serves them,and beamforming is realized through the coordination among BSs.In this thesis,transmit beamforming at BS and phase shifts at IRS are jointly optimized to maximize the minimum achievable rate/coverage probability to ensure the fairness of cell-edge users with the CSI bounded/statistical error model,and proposes an AO algorithm based on penalty method.Simulation results demonstrate that the proposed robust beamforming scheme can converge rapidly and outperforms reference schemes.Moreover,this thesis compares the proposed scheme based on the two models through simulation,where the bounded error model can guarantee absolute robustness,while the statistical error model has better performance in complexity.