Research On Key Technologies Of 5G Large-scale Distributed MIMO Downlink Based On GPP

The development of mobile communication technology has brought great convenience to mankind,but also further stimulated people's demand for mobile communications.Because of its large system capacity and high spectrum efficiency,large-scale distributed MIMO technology has gained widespread attention in academia.This paper mainly focuses on the downlink key technologies of large-scale distributed MIMO system,investigates the impact of the propagation delays among different users and different remote radio units on the performance of multiuser distributed MIMO systems,and studies the multiuser precoding scheme for distributed MIMO-OFDM with the consideration of the sample time offset.Finally,We present the implementation of downlink multiuser large-scale distributed MIMO system based on general-purpose processor.In the first chapter,we introduces the research background of the subject and the development process of mobile communication technology,then briefly introduces OFDM technology and the evolution of MIMO technology.Next,we give the research status of channel capacity analysis and precoding technology.In the second chapter,the basic architecture of the 5G experimental system and the data transmission flow of uplink and downlink are introduced,and the physical layer parameters and frame structure are given.Then some key technologies involved in the downlink of large-scale distributed MIMO system are briefly introduced,including the design of reference signal,precoding algorithms of BS and detection algorithms of UE.In the third chapter,we elaborates the channel model of large-scale distributed MIMO-OFDM system.According to the analysis,we obtain the conclusion that the propagation delays among different users and different RAUs will make the distributed MIMO-OFDM systems face more serious frequency selective channels,and give the mathematical model of distributed MIMO-OFDM with the consideration of the propagation delays.Then we analyze the channel capacity of the two models under MRT precoding scheme,and discuss the effect of antenna number,large-scale fading factor and propagation delays on system capacity through simulation.In the fourth chapter,we propose a low-complex multiuser precoding scheme with the consideration of propagation delays among different users and different RAUs.First of all,we introduce a precoding scheme that can completely eliminate the effect of propagation delays in the two-user model.Then,based on the least-square method,we obtain the universal precoding coefficient correction scheme in multiuser channel model.According to simulation,we analyze the influence of the subcarrier number and the maximum sampling offset on the system BER.It is pointed that the precoding coefficient correction scheme is feasible according to comparing the system performance with and without the precoding coefficient correction.In the fifth chapter,we discuss software implementation of downlink multiuser transmission algorithm for 5G-oriented large-scale distributed MIMO experimental system and present the hardware and software resources to implement the system.Because of the 5G experimental system's strict requirements of realtime performance,several methods for improving the software execution efficiency adopted in the software system are presented.The software implementation of downlink multiuser precoding algorithm at BS and detection algorithm at UE is given.Finally,we present the experiment result of the actual air-interface wireless transmission system.