Geometry-Based Stochastic Modeling for Non-Stationary High-Speed Train MIMO Channels

In this paper, a non-stationary geometry-based stochastic model (GBSM) for high-speed train (HST) MIMO channels is proposed. The proposed model employs geometry-based elliptical scattering model, where the received signal is a superposition of line-of-sight (LOS) and single- bounced rays. The time-varying reference system is introduced to more accurately characterize the non-stationarity of HST MIMO channels caused by the high speed factor. Based on the proposed model, the 2D space cross- correlation function (CCF) and the temporal autocorrelation function (ACF) are derived, simulated under both of isotropic and non-isotropic scattering conditions, and discussed in detail.