Research On Pilot Contamination Problem In Uplink Massive MIMO Systems

Massive MIMO is one of the key technologies of next generation communication.It mainly improves the performance of communication system by increasing the number of base station antennas.Massive MIMO has many advantages over existing communication technology,but it has a more serious bottleneck problem than conventional technology: pilot contamination.Pilot contamination severely constrains the performance of Massive MIMO systems.Therefore,this paper focuses on the pilot contamination in Massive MIMO systems.Because the TDD model of Massive MIMO system has the characteristic of reciprocity between uplink and downlink,this paper mainly studies the uplink pilot contamination problem.In this paper,the causes of pilot contamination and its impact on uplink performance are analyzed in detail,then the existing channel estimation method which can mitigate pilot contamination is studied.A pilot design method which can mitigate the pilot contamination is proposed,and an improved method based on the existing pilot assignment is proposed.The main research contents and research results are as listed below.1.The causes of pilot contamination in Massive MIMO systems and their impact on uplink performance are studied.First of all,a comprehensive analysis of the causes of pilot contamination is analysed.In addition to pilot multiplexing,this paper points out that hardware defects and non-complete reciprocity between uplink and downlink can also cause pilot contamination.Then comparison simulation of uplink performance with and without pilot contamination is presented in order to analyze the effect of pilot contamination on the uplink performance in Massive MIMO systems.2.The existing channel estimation method which can mitigate pilot contamination is analysed.Firstly,a channel estimation method which combines compressive sensing is studied.This method uses the spatial degrees of freedom caused by large number of base station antennas,and transforms the solution of the channel state information into the quadratic semi-definite programming problem.Then an approximate MMSE channel estimation method is studied,which can greatly reduce the complexity and maintain the same estimation accuracy as MMSE channel estimation method.3.A pilot design method and an improved pilot assignment method which can reduce the pilot contamination are proposed.Firstly,a pilot design method is proposed according to the time domain characteristic of pilot contamination.This method can estimate the channel state information by dividing the pilot transmission time and the pilot segment,and construct a new receiving matrix to effectively reduce the influence of pilot contamination.Then an improved pilot assignment method based on the existing pilot assignment method is proposed.This method sorts the users in the order of their channel states and their interference to target cell.Then an approximate MMSE channel estimation algorithm is used to get CSI.The simulation results show that the proposed methods can improve the accuracy of channel estimation compared with the traditional pilot design and pilot assignment method.