Incorporating diffuse scattering in geometry-based stochastic MIMO channel models

This paper introduces a model for incorporating the influence of the diffuse scattering in the MIMO radio propagation channel to the popular geometry-based stochastic channel models (GSCM). The proposed model is based on a superposition of clusters, each of which contributes to the channel as a multivariate random variable having a Kronecker structured covariance matrix. These covariance matrices are derived based on parametric modeling of the angular and delay power spectral densities related to each cluster. The model also includes polarization and temporal evolution of the clusters. The proposed approach is validated using a spectral-based visual analysis comparing the model with measured channel sounding data.