Design of Measurement-Based Stochastic Wideband MIMO Channel Simulators

In this paper, we propose a method for designing measurement-based stochastic models for simulating wideband multiple-input multiple-output (MIMO) wireless channels. The proposed method allows synthesizing channel transfer functions with the spatial-temporal-frequency correlation properties approximating those of real-world radio channels. The parameters of a channel simulator are determined by fitting the space-timefrequency cross-correlation matrix of the simulation model to the estimated space-time-frequency cross-correlation matrix of a physical channel. For this purpose, an iterative parameter computation algorithm has been developed. The proposed design method is verified by creating stochastic channel simulators based on the time-variant frequency responses (TVFRs) of real-world channels measured in different propagation environments.