Performance Analysis And Optimization Of GSIC-Based Cell-Free Massive MIMO-NOMA Systems

With rapid development of mobile communication,the demand for communication services has been continuously increasing.Huge traffic and massive connectivity have become important characteristics of 6G communication network.On one hand,compared to centralized massive multiple-input multiple-output(MIMO),cell-free massive MIMO has stronger anti-shadowing fading capabilities and can effectively overcome the edge effects and frequent base station switching problems in cellular network architectures.On the other hand,compared to orthogonal multiple access(OMA)technology,non-orthogonal multiple access(NOMA)has higher spectral efficiency,flexible support for larger numbers of connections,and is a key technology in the next generation of mobile communications.The combination of cell-free massive MIMO and NOMA can meet the future 6G communication demands for huge traffic and massive connectivity.This thesis explores the deep integration between cell-free massive MIMO systems and NOMA.NOMA based on group successive interference cancellation(GSIC)is applied to the downlink cell-free massive MIMO systems and a new architecture for cell-free massive MIMO-NOMA based on GSIC is introduced.Furthermore,a novel grouping method according to user’s equivalent path loss is developed to divide users into different groups,and the same pilot resource is reused among different groups,while users within the group use orthogonal pilots.The expression of the achievable user rate of the cell-free massive MIMO-NOMA system based on GSIC is derived,considering interference from other users within the group,interference from undemodulated groups,and interference caused by imperfect GSIC.The thesis compares the proposed approach with the traditional cell-free massive MIMO-NOMA system based on successive interference cancellation(SIC)and analyzes the advantages of the GSIC-based cell-free massive MIMO-NOMA system.To better reflect the channel characteristics and obtain more diversity gains,the thesis uses a Rician channel model with a line-of-sight component,and extends the access point(AP)from a single antenna to multiple antennas.Channels between AP antennas and the user is correlated.Furthermore,the thesis proposes a pilot allocation algorithm based on channel similarity to allocate pilots to group users,minimizing the maximum correlation coefficient among users using the same pilot among different groups to reduce pilot contamination and better ensure user fairness.The thesis studies cell-free massive MIMO-NOMA system based on GSIC under Rician channel model,and derives the expression of the achievable user rate,the line-of-sight component and multipath scattering non-line-of-sight component.The simulation results show that the GSIC-NOMA system with pilot allocation based on channel similarity has significant advantages in terms of system throughput and user rate compared to the GSIC-NOMA system with random pilot allocation.