The Research On Pilot Reuse Of Multi-User Distributed Antenna System

The fifth-generation mobile communications system (5G) will provide higher data rate, wider coverage and more stable communication links communication world for us, which requires higher spectrum efficiency and energy efficiency. Large-scale multi-input multi-output technology (MIMO) is one of the key technolo-gies of 5G. By increasing the number of antennas, MIMO can greatly improve the spatial resolution and increase the channel capacity. Large-scale distributed antenna system (DAS) is a kind of antenna deploy-ment of large-scale MIMO. By deploying antennas discretely in the cell, spectrum efficiency can be further increased. However, due to the conflict between the huge number of users in the cell and the limited number of orthogonal pilots, pilot contamination has become a key bottleneck in system capacity increase. One of the research topic in this paper is the spectrum efficiency of large-scale DAS considering pilot contamination. Another topic is the pilot allocation algorithm with low complexity of large-scale DAS.The paper firstly introduces the research background, reviews the development of large-scale MIMO technology, analyzes the characteristics, advantages and problems of large-scale DAS which leads to the research content of this paper.Then the paper describes the system model of a single cell large-scale DAS. We derive the system ca-pacity of the system with the ideal channel state information(CSI) based on the system model. By simulation, we compare the system capacity without pilot contamination and with the most serious pilot contamination. The huge gap of the simulation result shows that the large-scale DAS is an interference limited system, the impact of pilot contamination is far more than the impact of noise on the system. Therefore, to reduce pilot contamination is essential for large-scale DAS.Then the paper focus on the pilot contamination problem of large-scale DAS. Firstly, we study the chan-nel estimation of system, then we derive the lower bound of system capacity. Then, the asymptotic perfor-mance of the system capacity is given theoretically through the law of large numbers with the infinite number of RAUs. Finally, we simplify the formula by Hadamard inequality and obtain the upper bound of system capacity. After that, the paper introduce the simulation methods and parameters. Through the simple random pilot allocation algorithm, simulation results show that the derived theoretical values of capacity of the upper bound and lower bound are very accurate.Finally, the pilot allocation algorithm of large-scale DAS is studied. The farther the distance between the users, the finer the pilot contamination is, based on this basic point, combining the coloring theory in graph theory, we transfer the pilot allocation problem into abstract graph coloring problem. (1) Firstly, we propose a sparsification pilot allocation algorithm. By setting a threshold, the pilot contamination matrix can be sparsed qualitatively, then we can color the graph to allocate pilots. (2) Then, in order to solve the uncertainty of threshold in the algorithm, we put forward the min-max pilot allocation algorithm which is to minimizing the maximum interference. (3) Finally, the maximizing of the ergodic sum-rate is introduced which is proved to be theoretically optimal. The algorithm exhaustively calculate the system capacity for each user and each pilot to select the pilot when the system capacity reaches peak, we analyses the performance and the complexity of the algorithms so one algorithm can be selected to assign pilots in certain situation. Moreover, the optimal number of orthogonal pilots is given by simulation in certain system configuration.