Research On Pilot Contamination Mitigation Technology In Multi-cell And Multi-user Massive MIMO System

With the rapid development of information technology,the need for wireless communication data services is increasing day by day,traditional Multiple-In Multiple-Out(MIMO)technology can not meet the needs of current users.Therefore,the fifth generation of mobile communication system came into being,and Massive MIMO technology becomes a key technology in 5G which has been a research focus.In a massive MIMO system,hundreds or even thousands of antennas are deployed at base station.The advantages of massive MIMO technology will be realized when the number of antennas approaches infinity.Mainly including: the channels between users in same cell are orthogonal;the uncorrelated noise will disappear in system and the energy used by base station for each bit can be arbitrarily small.However,because of the same set of orthogonal pilots reused in all cells,pilot contamination can disturb channel estimation and lead to degrade the performance in system.Aiming at problem of pilot contamination,thesis uses pilot design technique and pilot scheduling technique to restrain the pilot contamination in multi-cell multi-user massive MIMO systems.The main work and innovations of thesis are summarized as follows:Firstly,thesis analysis the basic model of massive MIMO system and current pilot contamination elimination schemes for multi-cell multi-user massive MIMO systems which divided into four major categories: pilot design schemes,pilot assignment schemes,channels estimation schemes and precoding schemes.The thesis discuss the reasons of using TDD duplex mode and deduces uplink and downlink transmission processes of system,also analyzing the causes of pilot contamination and its impact on system performance.Secondly,the thesis solves the problem of pilot contamination using the pilotdesign technique,a pilot design scheme based on pseudo-random code assignment(PRAC)is proposed with allocation method and pseudo-random code technology.According to the interference intensity of the neighboring cells to target cell and the cross-correlation value between pseudo-random codes which have different time delays,the pseudo-random codes is assigned to the corresponding cells according to allocation method,then the pilot is designed by using corresponding pseudo-random code.Finally deducing the system performance indicators which obtained by the PRCA algorithm.Simulation experiments show the PRCA algorithm can enhance the cross-correlation between cells and reduce the impact of pilot contamination on system performance which improve the downlink rate of the users with the highest interference and achieve the balance the performance of mobile users.But the increasing of performance is at the expense of computational complexity.Finally,thesis solves the problem of pilot contamination using pilot scheduling scheme.A pilot scheduling algorithm based on cell classification and cross-entropy(CC-CE)is proposed with cell classification and cross-entropy technology.According to cell classification method,two-dimensional networks are divided into two types of one-dimensional networks,and two sets of orthogonal pilots are assigned to them which can reduce the number of interference cells.Next,introducing the basic theory of cross-entropy method and optimization problems.Based on cell classification,using cross-entropy method to solve pilot scheduling problem can further reduce the impact of pilot contamination on system.Simulation experiments show that the system performance obtained by the CC-CE algorithm is close to 99% of the system performance obtained by greedy algorithm.The computational complexity of CC-CE algorithm is only 1.23% of the greedy algorithm.Under the premise of ensuring performance of system,the computational complexity is greatly reduced.