Research On Mitigating The Influenceof Pilot Contamination In Massive MIMO Systems

The Massive MIMO(Massive Multi-Input Multi-Output), recently, is the emerging topic in communication system, which is the development based on the convetional MIMO(Multi-Input Multi-Output). One of the most significant merits is to increase the system capacity substantially.It is important to note that, the hypothesis about the wonderful performance of Massive MIMO could come true only when the system gets the perfect system channel statement information(CSI). However, it is quite difficult to the abtain the accurate CSI in the practical system because of the pilot contamination in Massive MIMO, which affects the system to achieve the expective performance. Reducing the impact of pilot contamination on Massive MIMO has become one of the biggest problem in Massive MIMO. Therefore, we put the attention to reduce the influence of pilot contamination on the system performance in Massive MIMO. Our main works in this paper are following.We analyze the technology of Massive MIMO and the pilot contamination in this system firstly, which provides the background knowledge for the further reaserch of this thesis.The second important part of our work is decreasing the pilot contamination influence on Massive MIMO performance, which could be presented by three ways: the channel eatimation, the matrix precoding method and the pilot assignment. The channel eatimation: we proposed a Fast-MMSE(Fast Minimum Mean Square Error,F-MMSE)estimation and angle of arrival limited Bayesian estimation(ALB). The F-MMSE is simpler and the ALB has stronger ability to resist the pilot contamination than the traditional MMSE estimation, respectively. The matri x precoding method: we presented a based MMSE multicell matrix precoding method(MMSE-MPC). Moreover, we present the closed form expression of matrix precoding method based on thorough theoretical derivation. The pilot assignment: we bring out a based Bayesian pilot assignment(BPA) that it could reduce the pilot contamination to great extent.