On Massive Mimo Performance With A Pilot Assignment Approach Based On Hungarian Method

In this paper, we propose a pilot assignment method to eliminate pilot contamination and improve the system performance in massive Multiple-Input Multiple-Output(MIMO) Time Division Duplexing(TDD) system. Cells are divided into center parts and edge parts according to the number of pilots and users. The pilot sequence sets are identical to center users, but mutually orthogonal for edge users in different cells. Hungarian method is illustrated to solve the pilot assignment in each pilot sequences subset. The simulation results show that the proposed approach achieves lower mean squared error(MSE) and higher sum rate(SR) in the system than the conventional design with a certain range of increasing signal noise ratio(SNR).