Precoding For Millimeter-wave Massive MIMO System

Today,with the rapid development of mobile communication,the demands of data rate becomes higher and higher,and people have begun to study for the next-generation mobile communication(5G)system.As one of the key technologies of the 5G system,Massive MIMO has become the focus of the current study,and the rich spectrum resources of millimeter wave band make it have more attention.Therefore,we study the precoding technology in millimeter wave Massive MIMO system.When using the digital precoding method,the system can achieve optimal performance,but considering the practical hardware cost,we focus on hybrid digital and analog hybrid precoding.1.We study for the massive MIMO system model and also give a model for millimeter wave massive MIMO system.Then we get the channel model using ray tracing method based on the finite scattering characteristics of millimeter wave.2.We study the digital and analog hybrid precoding technology in the single user millimeter wave massive MIMO scenario.The hybrid precoding based on orthogonal matching pursuit(OMP)algorithm is a classic solution,but it need to estimate the AOA and AOD of channel and its computational complexity is high.So we propose a hybrid precoding based on residual matrix iteration,which can reduce the complexity of system and get a good performance.3.we study the digital and analog hybrid precoding algorithm in the multiuser millimeter wave massive MIMO scenario.The PZF hybrid precoding is a scheme to solute the problem in this scenario,but it limits the number of the users must be the same as the number of RF chains,and its performance is not good.So we propose a digital and analog hybrid precoding algorithm that has more flexible and better performance.