| Mobile communication has brought great convenience for human society,at the same time,people’s requirements for mobile communications are increased.Driven by IOT(Internet Of Things)and MI(Mobile Internet),the fifth-generation mobile communication has given us new technical challenges.As one of the key technologies in the fifth-generation mobile communication system,the massive distributed MIMO tech-nology has attracted the attention of many scholars because it can make full use of the spatia-domain resources to improve spectrum efficiency and power efficiency.In the massive distributed MIMO system,the computa-tional complexity of the traditional block diagonalization pre-coding algorithm is high due to a large number of antennas at base station and user equipments.As a result,block diagonalization algorithm is difficult to be realized in the project.Therefore,the low-complexity multi-user pre-coding and interference suppression algorithm is one of hot spots in massive distributed MIMO system,and it is also one of focus of the thesis.In this thesis,a multi-user massive distributed MIMO system is used as a model to study the low-complexity pre-coding and interference suppression algorithms.Finally,the optimization of the 5G test system based on general-purpose processor(GPP)is given.Firstly,the thesis introduces the research background,describes the development process of MIMO technology,and points out that the massive distributed MIMO system will be the new architecture of the future mobile communication system.Multiuser pre-coding and interference suppression technology,as the key technology,is the focus of the thesis.Then the status of the current multiuser pre-coding and interference suppression technology is given.Secondly,this thesis describes the advantages of applying SDR to mobile communication systems,our 5G test system is based on GPP-based SDR platform.Then the thesis introduces our laboratory’s 5G test system in detail,mainly descirbes the basic architecture of the system,various configuration parameters,uplink and downlink data transmission process and frame structure.Thirdly,the thesis studies the multiuser pre-coding and interference suppression scheme in massive dis-tributed MIMO system.In this thesis,we analyze the massive distributed MIMO system model.Traditional block diagonalization(BD)pre-coding scheme ’s complexity is high,two improved low-complexity schemes are proposed based on BD scheme.The improved schemes utilize the TDD reciprocity of the channel,use GZI algorithm and GMI algorithm respectively to complete the unified design of downlink multiuser pre-coding and uplink interference suppression.The improved algorithms can greatly reduce the computational com-plexity compared with the traditional BD algorithm when the matrix dimension is large.At last,we compare the two algorithms with simulation.The result shows that,in the same case,GMI algorithm has larger system capacity and lower bit error rate compared with GZI algorithm.Finally,we optimize the implementation of 5G test system based on GPP platform because origin 5G test system ran slowly.First of all,MKL library helps to complete various matrix operations.It is not only convenient,but also can further improve the calculation speed,and our system can be upgraded automatically when MKL library is upgraded.Secondly,making full use of the Intel ICC compiler automatic optimiza-tion options,selecting the best compiler optimization option according to the specific scene can help us get considerable performance improvement.Finally,also one of the key points of this thesis,using multi-core and multi-threading technology for parallel computing with GPP’s multi-core hardware resources can greatly improve the system execution speed.Our 5G test system meet the real-time requirement due to the above optimization methods. |
PDF Full Text Request