Non-stationary wideband MIMO channel models for high-speed train wireless communication systems

Summary form only given. In high-speed train wireless communication systems, the underlying channels are essentially non-stationary due to fast-varying channel conditions. For the design, performance evaluation, and optimization of high-speed train wireless communication systems, it is a prerequisite to develop non-stationary channel models with time-varying small-scale fading parameters. In this talk, we propose novel non-stationary geometry-based stochastic models (GBSMs) with time-varying parameters for wideband high-speed train multiple-input multiple-output (MIMO) channels. The statistical properties of the proposed GBSMs are studied, such as time-variant spatial cross-correlation function and stationary interval. We also investigate the performance of spatial modulation systems employing the proposed GBSM. It has been demonstrated that the proposed channel models have the capability to characterize the non-stationarity of high-speed train channels and the stationary interval can well fit relevant channel measurements. The talk will end with discussing recent advances and future challenges in MIMO channel models.