| Due to large number of antennas deployed at the base station,massive MIMO technology could significantly improve the spectrum efficiency and energy efficiency,which became one of the key technologies in fifth generation mobile communication system.However,pilot contamination caused by the pilot multiplexing among cells has become a major factor restricting the improvement of system performance.Reducing pilot contamination is of vital importance to the current massive MIMO research.This thesis studies the strategies of reducing pilot contamination from the perspective of pilot allocation,and focuses on the issue of pilot allocation based on user classification and user location information.The specific research contents are as follows:1.Aiming at improving accuracy of traditional user classification methods and the poor system capacity of the traditional pilot allocation algorithms,this thesis proposes a pilot allocation algorithm based on user classification.Firstly,the distribution of all users in the cell is determined according to the mean and variance of signal strength of the users in the cell.Secondly,users are divided into center users and edge users by setting dynamic signal strength thresholds based on the distribution of users in the cell.At last,allocating pilots to center users by using the improved maximum carrier-to-interference ratio algorithm due to the good channel quality and high robustness under pilot contamination of the center users.while allocating pilots to edge users using the greedy algorithm owing to the poor channel quality and poor robustness under pilot contamination of the edge users.Results show that the proposed user classification algorithm has higher accuracy.The proposed pilot allocation algorithm can significantly improve the system capacity and reduce the negative impact of pilot contamination while ensuring the communication performance of the edge users.2.Aiming at the problem that the existing algorithms were mostly complicated and the performance gap among users was large,a pilot allocation algorithm based on user location is proposed.Firstly,under the power controlling of system,derive the user uplink signal to interference and noise ratio expression in detail.Secondly,using the ratio of the distance between the user and the base station of target cell to the distance between the base station of the interfering cell to measure the pilot contamination intensity of the interfering cell caused by the user.At last,according to the pilot contamination intensity among the users in different cells,using pilot contamination intensity balanced algorithm to allocate pilots to users,namely,the same pilot is allocated to two users with less pilot contamination.Results show that the proposed algorithm can reduce the performance gap between users with lower complexity and reduce the adverse impacts of pilot contamination on the system performance in a more practical way. |
