| Massive MIMO has increased the number of base station antennas from a few,a dozen to several hundred,and gained the advantages of improving spectrum efficiency and reducing delay and noise interference,thus the Massive MIMO adopted by the fifth generation communication system as one of its key technologies.However,because of the limitation of the channel coherence time,Massive MIMO cannot allocate pilot sequences that are orthogonal to each other in the system,which leads to pilot contamination and affects the accuracy of channel state information estimation.And the pilot contamination makes a bottleneck which the development of this technology encountered.Therefore,exploring and studying pilot decontamination methods in Massive MIMO system has great significance for improving the performance of communication system.This paper mainly studies the uplink transmission part of Massive MIMO system.From the perspective of improving pilot allocation scheme,optimizing pilot sequence design and uplink channel estimation algorithm,two sets of pilot decontamination schemes are proposed which solves the negative impact on Massive MIMO system.First,this paper analyzes the system model and signal transmission scheme of Massive MIMO,and the cause of pilot contamination in Massive MIMO system is theoretically derived and analyzed.The analysis results show that pilot contamination has brought negative effects to the system,such as user signal to interference plus noise ratio reduction,system capacity reduction.Secondly,to deal with the problem of pilot contamination caused by the reuse of the same pilot sequence in each cell in a Massive MIMO system,this paper proposes a scheme that combines polar pilot assignment and pseudo-random code pilot design.First,all users in each cell were marked with polar coordinates,and pilot sequences were allocated according to the size of the user's polar Angle.Then,pseudo-random codes with different time delays were used to disturb to users using the same pilot sequence in different cells.In this way,the adverse impact of pilot contamination on channel estimation is reduced.Compared with other traditional schemes,this scheme can reduce the normalize mean square error of system channel estimation by about 1.43-2.83% on the basis of other schemes.The cumulative distribution function performance of the signal to interference plus noise ratio is improved by 28%,and the user transmission rate is increased by 40%.Close to the performance of the system without pilot contamination.At the same time,aiming at the problem that the channel estimation has errors due to the pilot contamination and the complexity of the traditional channel estimation algorithm is large,which affects the performance of the system,a algorithm based on the angle-domain and SVD(Singular Value Decomposition)in the Massive MIMO system is proposed in this paper.Taking the advantage of the feature which the spatial correlation channel of the Massive MIMO system has the advantage of approximate sparsity in the angle-domain,to perform channel estimation.And for reducing the complexity of channel estimation algorithm,the algorithm replaces the process of channel autocorrelation matrix inversion in channel estimation by SVD.Compared with the traditional channel estimation scheme,this scheme can effectively improve the normalize mean square error performance of the channel estimation by 70-85%,and increase the average transmission rate of the user uplink transmission by 36%.Finally,this paper uses the new scheme proposed in this paper to jointly deal with the problem of pilot contamination in Massive MIMO system.And perform performance simulation on it.Through research,it can be concluded that compared to the two schemes proposed in this paper,the combined scheme improves the normalize mean square error performance of channel estimation by about 8-14.8%,and the cumulative distribution function performance by 2.3-7.3% The average transmission rate of uplink transmission is increased by about 3-6.6%. |
PDF Full Text Request