Research On Millimeter Wave Massive MIMO Channel Estimation And Hybrid Beamforming

In such an era of Internet of things,the types and number of intelligent terminals increase boomingly.The fourth-generation communication system(4G)can't afford the growing demand of communication services.The massive MIMO wireless communication technology,which is widely used in millimeter wave(mm Wave)system,is one of the three key technologies in 5G,and has become the first problem to be overcome.Different from the conventional MIMO system,massive MIMO is equipped with an antenna array at the base station,in order to improve the data transmission speed,data transmission quality,energy utilization,spectrum utilization,communication system capacity and so on.With the expansion of antenna array scale,the existing channel estimation algorithms and beamforming structures are no longer applicable.How to use the characteristics of millimeter wave massive MIMO system effectively,achieve lower complexity and higher accuracy channel estimation,improve the performance of beamforming,and reduce the energy consumption,cost and complexity of beamforming,has become the key to implementation of millimeter wave massive MIMO technology.The channel estimation of mm Wave massive MIMO is the basis of beamforming technology.The traditional digital beamforming and analog beamforming are not suitable for mm Wave massive MIMO system.Instead,the hybrid digital-analog beamforming technology is used.However,it still has the disadvantages of high complexity,high energy consumption and high equipment cost.In order to effectively reduce the energy consumption and cost of the system,it's important to decrease the number of RF chains and optimize the structure of analog beamforming circuit.In this paper,a channel estimation method based on adaptive Stagewise orthogonal matching pursuit is proposed.Firstly,several existing signal reconstruction methods for sparse channels are described including signal reconstruction methods based on OMP,Co Sa MP and St OMP.The principle of the above algorithm is analyzed in detail,and then an improved channel estimation method based on ASt OMP is proposed.Specifically,the sparsity of the signal is estimated before the iteration.During the iteration,the number of selected atoms is adaptively adjusted according to the iteration progress,and the atoms with poor overall correlation in the atomic set are deleted,so as to complete the signal reconstruction quickly and accurately.The simulation results verify that the operation efficiency and the success probability of signal reconstruction are improved,and the normalized mean square error of signal reconstruction is reduced.In this paper,the structure of single RF link in hybrid beamforming is studied.Firstly,several classical digital beamforming technologies are analyzed in detail,including ZF based,MMSE based and BD based digital beamforming technologies.The principle of algorithms and the reasons why they are no longer applicable in mm Wave massive MIMO system are described.The hybrid beamforming and the improved single RF link structure are described followed.Then,several improved methods for single RF link hybrid beamforming structure are introduced,including hybrid beamforming based on phase shifter networks and hybrid beamforming based on switchs and inverters.Compared with the above methods,a single RF link hybrid beamforming structure based on fixed phase shifter and amplifier is proposed.The simulation results show that the sum rate and energy efficiency of the system are improved.