NOMA-based Resource Allocation In Intelligent Reflecting Surface-assisted Cloud Radio Access Networks

Application scenarios of 6G will place higher demands on performance metrics such as bandwidth,connectivity and coverage.Millimeter wave provides abundant available bandwidth for 6G,while suffer from severe path loss.Beamforming technology and network densification under Cloud Radio Access Networks(C-RAN)provide effective solutions to combat millimeter wave signal fading.On the other hand,non-orthogonal multiple access(NOMA)technology can support multi-user parallel access in a single beam through rational resource allocation,enabling the massive connectivity needs of 6G network devices.Intelligent Reflective Surface(IRS)can intelligently change the radio channel through reflective beamforming design to achieve signal enhancement,and enhance network coverage.Therefore,thesis introduces IRS into NOMA-based millimeter-wave C-RAN systems,and investigates the joint beamforming and resource allocation algorithms for IRS-assisted NOMA systems under single-cell and multi-cell scenarios.For the single-cell scenario,the IRS is used to overcome signal blocking and path loss of millimeter-wave NOMA communication,moreover,the impact of intra-group user interference caused by NOMA can be mitigated by rational design of user grouping,beamforming,and power allocation.In addition,this thesis considers the low-cost requirement of C-RAN architecture,the low-cost hybrid beamforming structure is deployed,and the phase shift of IRS reflection elements and analog beamforming is quantified.A low-complexity non-iterative optimization algorithm is proposed to address the design challenges of jointly optimized user grouping,beamforming and power allocation algorithms due to the introduction of IRS.Specifically,the reflective beamforming is designed by Riemann optimization algorithm to maximize the user channel gain,then,a dynamic user grouping algorithm based on the singular value decomposition(SVD)of the equivalent channel is proposed,finally,the power allocation optimization algorithm is designed based on the minimum mean square error property.The results show that the proposed scheme can effectively improve the system spectral efficiency compared with the existing IRS-assisted NOMA using 2-user grouping scheme,NOMA without IRS scheme or IRS-assisted orthogonal multiple access schemes.For the multi-cell scenario,this thesis mainly addresses the serious inter-cell interference and intra-group interference caused by NOMA for cell edge users.By exploiting the centralized resource allocation and interference management capabilities of the C-RAN architecture,and deploying IRS at the cell edge to change the radio channel,then,effective interference suppression is achieved by jointly optimizing reflective beamforming of the IRS,transmit beamforming of the baseband unit(BBU)and power allocation of remote radio heads(RRHs).Due to the joint optimization problem is nonconvex,an alternating optimization based joint optimization algorithm is proposed.Specifically,the non-convex power allocation and transmit beamforming optimization problems are transformed into three low-complexity convex subproblems by the minimum mean square error detection,and then solved by iteration;the reflective beamforming optimization algorithm is designed by applying successive convex approximation and semi-definite relaxation algorithm.The research results verify the superiority of the proposed scheme in terms of multi-user access and user interference suppression.