Research On CSI Feedback And Antenna Selection In Large Scale MU-MIMO Systems

In recent years,the types of data services have become more abundant.In order to pursue better user experience and carry increasing traffic and users,the industry has started research on the fifth generation(5th Generation)mobile communication system.Massive Multi-user Multiple Input Multiple Output(MU-MIMO)which deployed large-scale antenna arrays at the base station can serve more users without increasing bandwidth and power.In massive MU-MIMO transmission,the terminals feed back channel state information(CSI)to the base station through limited feedback in order to adapt to the channel conditions and ensure high-reliability and high-speed communication.Then the base station selects the appropriate antenna group for data transmission through the antenna selection technology,reducing the hardware cost while obtaining higher spectral efficiency.However,with the increase of the number of antennas,the feedback of CSI channel state information sharply increases,and the complexity of the joint receiving and transmitting antenna selection technology is also increased.In this paper,we study how to reduce the feedback of CSI channel state information in massive MU-MIMO system under FDD transmission mode.Considering the space-time correlation of massive MU-MIMO system channels,we use tensor data structure to describe CSI channel state information from multiple dimensions and construct a multi-dimensional CSI channel state information tensor.And then we use tensor decomposition algorithm to predict missing CSI channel state information.Second,we use the suboptimal maximization of the channel gain target to join the receiving and sending antenna selection.So the maximized channel gain by joining receiving and transmitting antenna selection is transformed into a typical optimal target matching problem,which is solved by using the Hungarian algorithm of low complexity and high efficiency.In addition,the proposed scheme about the limited feedback of CSI channel state information based on tensor decomposition prediction and the joint receiving and transmitting antenna based on Hungarian algorithm both pass the simulation experiments,and verify the validity of the two schemes.