Abstract
For high-resolution massive MIMO and very large antenna arrays, wireless channel models have to scrutinize the detailed space features of the surrounding environment. Existing models such as WINNER and 3GPP are not appropriate for validating and evaluating new concepts for 4G/5G as they do not consider the spatial characteristics of the real environment. The simplified 3D shapes of simulated objects in geometry-based channel models, which are constructed using vertical and horizontal planes, may cause significant difference from the real channel. In this paper, we present an approach to model the specular reflection of a signal from an arbitrary inclined surface by taking into account the signal's polarization and a spatial distribution of massive MIMO antenna elements. The approach was validated through simulating LTE uplink transmissions in an environment modeled based on Google Maps. Results showed the importance of considering detailed 3D characteristics of the surroundings in simulations. We observed that even slightly inclined walls can have significant influence on channels in comparison with models with only vertical and horizontal surfaces due to different propagation paths, different angles of reflection, and different changes of polarizations.