Codebook Design And Optimization For Massive MIMO Systems

In recent decades,communication technology has developed rapidly,especially the launch of smart phones represented by iPhone and Android in 2007.People have entered the era of "Mobile Internet","IoT"(Internet of Things)and "IoE"(Internet of Everything)will come one after another,and wireless communication technology is the premise of all those.Therefore,5G has long been put on the charter.Due to the advantages of Massive MIMO technology in terms of spectrum efficiency,delay time and transmission rate,it has become one of the core of 5G technologies.The antenna size of Massive MIMO system is far less than that of conventional MIMO.For this reason,its requirements for precoding performance are far superior to traditional MIMO.The huge size of the antenna array means that a large number of matrix operations,research and proposed low-complexity coding precoding algorithms have become the top priority.Based on this,this thesis studies the precoding algorithm in Massive MIMO,and focuses on the codebook-based precoding algorithm.The specific work is as follows:1.Introduce the development background and current situation of 5G and Massive MIMO systems,analyzes the differences with traditional MIMO,and the advantages in various aspects,introduces and analyzes the system capacity and antenna array.2.The precoding algorithm is studied.Several classical precoding algorithms in MU-MIMO system are introduced.After that,the precoding algorithm based on codebook is introduced in detail and simulated.3.The performance of channel capacity and bit error rate of Massive MLMO system under unitary array perturbation is analyzed,and Kerdock codebook is improved and optimized according to the change,and its performance is verified by simulation.4.According to the improvement principle of Hill sort to Selecting sort,the traversal and feedback of the rotating DFT codebook are improved,and the codeword selection is improved,too.The simulation results show that it has better performance in terms of bit error rate and spectrum efficiency than the original rotating DFT codebook.