Performance Optimization Of Cell Free Massive MIMO-NOMA Systems

With the development of technology,wireless communication networks need to have better performance to support future emerging applications.As one of the popular technologies in recent years,cell free massive Multiple-Input Multiple-Output(MIMO)has attracted a lot of attention from researchers.It can effectively avoid problems such as frequent handover,severe interference,and poor communication quality for edge users in cells.Non-Orthogonal Multiple Access(NOMA)is considered as a key multiple access technology that can provide better spectrum efficiency,system capacity,and user fairness for communication systems.In future research on the sixth generation of mobile communication systems,the combination of these two technologies is expected to solve cell interference problems while improving system spectrum efficiency to meet the communication performance requirements of emerging applications in the future.Therefore,this thesis will study the cell free massive MIMO-NOMA system.The main research contents are as follows:Firstly,this thesis mainly analyzes the uplink and downlink communication performance of the cell free massive MIMO-NOMA system,and derives the relevant equations for signal transmission in the communication links.In the uplink communication,the maximum ratio combining technology is used and the impact of imperfect Successive Interference Cancellation(SIC)is considered.Based on the channel statistical characteristics,the closed-form expression of the system’s Sum Spectral Efficiency(SSE)is derived.Similarly,in the downlink communication,the closed-form expressions of the system’s SSE are derived under the imperfect SIC,using both the traditional maximum ratio precoding and the full-pilot zero-forcing precoding with the same forward overhead.Then,the user clustering issue when using NOMA technology is discussed and analyzed.Starting from the existing user clustering algorithms,a new user clustering solution is proposed.In addition,the Mean-shift algorithm from machine learning is employed to improve the clustering algorithm and obtain more stable clustering results.The simulation results demonstrate that the user clustering algorithm proposed in this thesis has a better effect on improving the performance of cell free massive MIMO-NOMA systems compared to other user clustering algorithms proposed in other literature.Finally,to further optimize the performance of the cell free massive MIMO-NOMA systems,this thesis proposes an optimization design for the uplink optimal backhaul link coefficient and downlink power allocation.Using the optimization design of the backhaul link coefficient can effectively reduce interference to the user signal.The optimization objective of downlink power allocation is to maximize the SSE.The successive convex approximation algorithm is employed,and the problem of downlink power optimization is transformed into a geometric programming problem.The feasibility of the proposed optimization methods is verified through simulation result analysis.