3D Massive MIMO Channel Modeling

Massive MIMO(Multiple-Input Multiple-Output)technology is one of the most potential 5G key wireless transmission technologies,which has demonstrated providing the greater system capacity and higher spectral efficiency and has attracted worldwide attention.It is crucial to explore a channel model that can effectively characterize the propagation characteristics of Massive MIMO channels.The channel model can be used to evaluate and optimize Massive MIMO techniques,algorithms,products and systems.The 3D channel modeling of Massive MIMO system is studied in this paper and the main content of the thesis is as follows:Firstly,the characteristics of multipath propagation and fading of wireless channel is investigated in this paper.The geometry-based stochastic channel modeling methods and correlation-based stochastic channel modeling methods which are widely used in MIMO channel modeling are summarized.Then,the propagation characteristics of Massive MIMO channel such as the near-field and non-stationary characteristics are studied.The near-field effect is modeled by spherical wavefront,and the second-order approximation of the spherical wavefront is studied,which simplifies the spherical wavefront as the parabolic wavefront,rather than the plane.The non-stationary characteristics on array axis show that any scattering cluster cannot affect the whole array.Massive MIMO channel cannot be regarded as stationarity.In addition,simulation shows that Massive MIMO channels tend to be orthogonal with increasing number of antennas.Secondly,the 3D MIMO channel modeling method based on 3GPP is studied.The path loss model of 3D-UMa(Urban Macrocell)and 3D-UMi(Urban Microcell)are analyzed.The simulation results show that the path loss is proportional to the carrier frequency.The path loss under the LOS condition is less than the NLOS condition,the higher the average height of the building has the higher the path loss.The spatial correlation of 3D MIMO channels is studied by different angle distribution models,the simulation results show that the angular distribution of Von Mises Fisher can provide greater channel capacity.In addition,a 3D Massive MIMO Kronecker channel model is proposed,which the non-stationary evolution of scattering clusters along the array axis is modeled using the birth-death process.Simulation results show that the proposed channel model can not only characterize the antenna correlation but also the evolution of scattering clusters on the large-scale antenna array axis.Finally,a non-stationary 3D spatial channel model based on stochastic scattering clusters for Massive MIMO systems is proposed.A parabolic front is used instead of a spherical wavefront,which reduces the computational complexity of the 3D channel model.In this paper.On basis of the distribution of scattering clusters in space,by introducing two physical concepts of effective probability of scattering cluster and effective scattering cluster,a simple modeling method based on spatial location information of scattering cluster is provided.The stochastic process is used to model the appearance and disappearance of effective scattering clusters in the array axis.In addition,the statistical properties of the model are deduced.The simulation results have verified the rationality of the proposed model.