On The Anti Pilot Contaminatin Channel Estimate And Precoding Design For Massive MIMO System

Known for its high spectral efficiency, the MIMO technology has become the core technology in LTE system after the rapid development in the last decade. With the ever increasing user demand for better spectrum efficiency and higher throughput, the massive MIMO system has received much attention and become one of the most potential techniques towards 5G technology. In the very large scale MIMO system (massive MIMO), the base station is assumed to be provisioned with many (hundreds to thousands) antenna, while providing services for relatively few users. With the increase in the number of antennas, there are some useful properties introduced. For example, now one can greatly improve the efficiency of spectrum utilization and energy efficiency. By increasing the number of antennas to infinity (or large enough), the performance of massive MIMO system will be dominated by the inter-cell interference due to the pilot contamination, wherein those adjacent cells share the same set of pilots, thus leading to the uplink channel estimation quality degraded, which will further constrain the achieved channel capacity, and the level of the pilot contamination will not be dependent on the antenna number, but the large scale fading factor.This paper will firstly review the basic system model of the massive MIMO system. And the achieved performance under ideal channel estimate for both the single cell and multiple cell provisioned with massive MIMO are presented. Then this paper discusses the uplink channel estimation problem for the massive MIMO system. As for the channel estimation in uplink of massive MIMO base station, the paper presents the simulation analysis of the direct estimation method and the MMSE algorithm in both the single cell and multi cell scenarios. And the simulation results show that, the pilot contamination will seriously degrade the channel estimate accuracy for both the direct estimation method and MMSE algorithm. And a Bayesian estimation algorithm is proposed in this paper to enable a cooperation among different base stations, and it is shown that the channel estimate can be improved by exploiting the related a-priori information involved, which provides some hints on how to further improve the channel estimate for massive MIMO system.Then we analysis the precoding performance in downlink.Derive the algorithm in one-cell system, simulate the performance of MF and ZF. Which is good in non-pilot contamination case but terrible in pilot contamination.we introduce a pilot contamination precoding algorithm which take the large scale efficient of the neighbor cells into consideration and eliminate the impact of the inter-cell interference.