Research On Downlink Transmission Performance Enhancement For Cell-free Massive MIMO Systems

In order to meet the higher development requirements of 6th-generation mobile communication(6G)in terms of spectral efficiency,energy efficiency,massive connection,delay and reliability,experts and scholars around the world have widely paid attention to cell-free massive multiple-input multipleoutput(MIMO)systems.Much research has been carried out on the uplink transmission of cell-free massive MIMO systems,but the research on the downlink transmission is not enough.Therefore,it is necessary to study the downlink transmission performance enhancement.In this thesis,due to the characteristics of high scalability,low cost and low power consumption,the problems of inadequate channel hardening,limited backhaul link resources,limited access point power and cost,and large inter-user interference in cellfree massive MIMO systems are studied.The downlink spectral efficiency of the systems is improved,and the downlink transmission performance of the systems is effectively enhanced.The main innovation points are as follows:Aiming at the problem of incomplete channel hardening in cell-free massive MIMO systems,a selective user downlink channel estimation strategy based on channel hardening difference is proposed.In the proposed strategy,a dynamic threshold model of channel hardening ratio is established,and users are divided into groups by using the threshold.A differentiated downlink channel estimation strategy for different user groups is designed,and the downlink spectral efficiency and pilot overhead of the proposed strategy are analyzed.Theoretical analysis and simulation results prove that the proposed strategy achieves downlink spectral efficiency close to that of all users downlink channel estimation strategy with no more than 60%downlink pilot overhead,and its downlink spectral efficiency is significantly better than that of no downlink channel estimation strategy and close to the case of known perfect channel state information.Moreover,the impact of the hardening ratio threshold on the downlink spectral efficiency and pilot overhead is simulated and analyzed,and the hardening ratio threshold is further optimized.To deal with the problems of limited resources of backhaul links and limited power and cost of access points in cell-free massive MIMO systems,an adaptive access point selection scheme based on simulated annealing algorithm is proposed by adopting the idea of "user-centric".The proposed access point selection scheme takes the association between the user and the optimal access point as the initial solution,the large-scale fading coefficient decreasing as the iterative order,and the expected downlink spectral efficiency as the optimization objective,the number of better large-scale fading coefficients as the chain length,and uses the dynamic iterative method to gradually obtain the optimal solution.Simulation results show that compared with large-scale fading based,K-Means based access point selection scheme and fully connected scheme,the proposed access point selection scheme significantly improves the downlink spectral efficiency of the system.The problems of limited resources of backhaul links,limited power and cost of access points and large inter-user interference in cell-free massive MIMO systems are discussed,and an improved partial pilot zero-forcing precoding scheme based on interference coefficient is proposed.Based on the classical partial pilot zero-forcing precoding,the proposed scheme introduces interference coefficient,and uses this coefficient to design a new user grouping scheme.The simulation results show that compared with the Maximum Ratio Transmission precoding scheme and the classical Partial Pilot Zero Forcing precoding scheme,the proposed improved partial pilot zero forcing precoding scheme based on interference coefficient improves the downlink spectral efficiency of the system to a certain extent.