Research On Multi-user Communication Methods Assisted By Intelligent Reflecting Surface

As a promising technology,intelligent reflecting surface(IRS)can improve communication performance by changing the wireless propagation environment.By adjusting the electromagnetic response of each IRS element,IRS is capable of reconfiguring the propagation of the wireless signal,thus realizing the concept of a smart radio environment.Instead of passively adapting to the dumb wireless environment,IRSassisted communication provides a new paradigm shift for wireless communication by actively changing the wireless environment.Therefore,IRS-assisted communication has become a key candidate for the nextgeneration communication system.Considering that the number of user devices to be connected in future networks will further increase rapidly,one of the most fundamental and critical issues in the integration of IRS into wireless communication systems is how to achieve efficient multi-user communication.Solving this problem is of great importance to promoting the development and implementation of IRS-assisted communication technologies.To this end,based on different multiple access technologies,this dissertation investigates the IRS-assisted multi-user communication methods from link-level to system-level communication scenarios by employing mathematical optimization theory,and proposes an IRS reflection beamforming method for achieving the multi-user rate boundary;proposes a non-orthogonal multiple access(NOMA)based joint transmission and reflection beamforming optimization method for maximizing the throughput;proposes a joint IRS deployment,IRS reflection beamforming,and resource allocation optimization method for maximizing the system weighted sum rate.The main contributions of this dissertation are summarized as follows.First,the rate boundary characterization problem in the IRS-assisted single-antenna multi-user communication is studied.In particular,the concept of dynamic IRS configuration is proposed,and the corresponding rate boundary characterization problems for IRS-NOMA and IRS-OMA(Orthogonal Multiple Access)are formulated by employing the rate-profile technique.Firstly,the ideal case is considered,where the time interval required for two adjacent IRS configurations is ignored.Based on the timesharing condition,the complete rate boundaries of IRS-NOMA and IRSOMA are characterized by using the Lagrangian dual method,which reveals the corresponding optimal IRS reflection beamforming policy.Then,the general case is considered,where the time interval required for two adjacent IRS configurations is non-negligible.Sub-optimal algorithms are proposed to find a rate inner bound for IRS-NOMA and IRS-OMA.Simulation results show that IRS can significantly enhance the rate gain for multi-user transmission,and the dynamic IRS configuration can further improve this rate gain.Second,the design of IRS-assisted multi-antenna multi-user NOMA communication is investigated.The employment of NOMA in IRS-assisted multi-antenna multi-user transmission can further improve communication performance and connection capability.The joint transmission and reflection beamforming optimization problem is formulated for maximization of the throughput,subject to the successive interference cancellation decoding rate constraint of multi-antenna NOMA and different types of IRS elements.An alternating optimization-based iterative algorithm is proposed.For solving the transmission beamforming optimization subproblem,SCA(successive convex approximation)and SDR(semi-definite relaxation)methods are invoked.For solving the IRS reflection beamforming optimization subproblems under three different types of IRS elements,SCA,SROCR(sequential rank one constraint relaxation),and relaxation-quantization methods are respectively employed.Simulation results show that the proposed algorithm can effectively improve the throughput in the considered IRS-assisted multiantenna multi-user NOMA communication,and the achieved performance significantly outperforms other benchmark schemes.Third,the fundamental IRS deployment design problem in IRSassisted multi-user communication systems is investigated.Considering different multiple access schemes,the joint optimization problem of IRS deployment location,IRS reflection beamforming,and user power allocation is formulated for the system weighted sum rate.Firstly,a monotonic optimization-based algorithm is developed to find the performance upper bound,in which the Polyblock outer approximation algorithm,SDR,and the exhaustive search method are employed.Secondly,in order to reduce the complexity,an alternating optimization-based iterative algorithm is developed to find a high-quality sub-optimal solution,in which the power allocation,IRS reflection beamforming,and IRS deployment location are respectively optimized by involving SCA,SROCR,and local region optimization methods.Simulation results show that the optimization of IRS deployment location can significantly improve the multi-user communication performance,and the proposed sub-optimal solution algorithm can achieve the near-optimal performance with low complexity.