In this paper, the conventional method of establishing spatial channel models (SCMs) based on measurements is extended by including clusters? of? scatterers (CoSs) that exist along propagation paths. The channel models resulted utilizing this new method are applicable for generating channel realizations of reasonable spatial consistency， which is required for designing techniques and systems of the fifth generation wireless communications. The scatterers’ locations are estimated from channel measurement data obtained using large? scale antenna arrays through the Space? Alternating Generalized Expectation? Maximization (SAGE) algorithm derived under a spherical wavefront assumption. The stochastic properties of CoSs extracted from real measurement data in an indoor environment are presented.
channel modeling; large?scale antenna array; spatial consistency; space?alternating generalized expectation?maximization (SAGE) algorithm; spherical wavefront