Research On Pilot Distribution And Channel Estimation Of Massive MIMO System

Massive MIMO(Multi-input Multi-output)has become one of the most important research directions in the field of communication.Its technology development and maturity are constantly promoting the fifth generation mobile communication business.Based on the Massive MIMO technology,a large number of antenna arrays can be configured on the base station side,and the stability of the wireless link and the data transmission rate can be greatly improved by using simple signal processing technology.However,the problem of pilot contamination and imperfect channel estimation leads to the failure to get accurate channel state information(CSI),and other issues,which has been restricting the application of Massive MIMO system.In view of the above problems,this thesis studies the scheme of improving pilot contamination and channel estimation performance of Massive MIMO system.The main work is as follows:First,the Massive MIMO system model is introduced,and the cause and influence of pilot contamination in Massive MIMO system are studied.The common solutions to solve the problem of pilot contamination are analyzed and summarized,which lays a foundation for the scheme to reduce pilot contamination.The influence of pilot contamination on the performance of commonly used channel estimation algorithms is compared by simulation.A pilot assignment algorithm is proposed to ensure user fairness in Massive MIMO system.For the edge users in the target cell,it can minimize the interference which is from the users adopting the same pilot in adjacent cells.The algorithm can balances the SINR of each user in the target cell and users who can avoid poor channel cannot meet the communication needs.The proposed algorithm ensures user fairness while reducing pilot contamination.Finally,this thesis makes a comparison and analysis of two common linear estimation algorithms and proposes an efficient and practical one that does not required knowledge of large-scale fading coefficients for the case when the communications channels present multipath fading behavior.The minimum variance unbiased estimation is used to calculate the combined interference value of thealgorithm,the original value of the MMSE channel estimation algorithm is replaced by the estimated value of the parameter.Theoretical and simulation results show that the proposed channel estimation method is efficient and practical with the increasing number of antennas,and its performance is close to the ideal MMSE channel estimation.