Novel 3-D Non-Stationary Wideband Models for Massive MIMO Channels

In this paper, a novel 3-D non-stationary wideband geometry-based stochastic theoretical channel model for massive multiple-input multiple-output communication systems is proposed. First, a second-order approximation to the spherical wavefront in space and time domains, i.e., parabolic wavefront, is proposed to efficiently model near-field effects. Second, environment evolution effects are modeled by spatial-temporal cluster (re) appearance and shadowing processes. We propose (re) appearance processes to model the visibility of clusters with enhanced spatial-temporal consistency. Shadowing processes are used to capture smooth spatial-temporal variations of the clusters' average power. In addition, a corresponding simulation model is derived along with a 3-D extension of the Riemann sum method for parameters computation. Key statistical properties of the proposed model, e.g., the spatial-temporal cross-correlation function, are derived and analyzed. Finally, we present numerical and simulation results showing an excellent agreement between the theoretical and simulation models and validating the proposed parameter computation method. The accuracy and flexibility of the proposed simulation model are demonstrated by comparing simulation results and measurements of the delay spread, slope of cluster power variations, and visibility regions' size.