Radio Wave Propagation Simulations Based On Point Clouds – Methods, Experimental Validations and Applications To Radio Link Design

The exponential increase in mobile data traffic has resulted in spectrum shortage at the sub-6 GHz frequencies. The fifth-generation (5G) of wireless technologies aims to solve this by utilizing the millimeter-wave (mm-wave) bands and denser base station deployment. Cellular coverage at mm-wave bands and the interference conditions of dense spatial re-use of frequency spectrum are not well known. To evaluate link performance and predict coverage, wave propagation at radio frequencies in various scenarios must be characterized. This thesis provides improved tools, methods, and insights for coverage simulations to plan future deployments of wireless networks.Radio channel measurements are essential to understanding propagation and coverage characteristics. This thesis presents insights obtained from radio channel measurements in emerging environments, the outdoor-to-indoor (O2I) case and an elevator shaft. In the former case, the role of scattering from outdoor and indoor features is evaluated at 4 GHz and 14 GHz. In the latter, an Universal Software Radio Peripheral-based channel sounder is developed for polarimetric wideband channel measurements in an elevator shaft at 2.45 GHz and 5.8 GHz. The results show that the elevator shaft is a polarization-selective environment, with attenuation differences of up to 20 dB between links polarized along different sides of the elevator shaft. In addition to measurements, site-specific propagation simulation is an important tool in understanding radio wave propagation and its effect on coverage. Simulation accuracy is affected by level of detail in model of the propagation environment. In this work, ray-tracing and ray-launching propagation simulation methods based on detailed laser-scanned point clouds are developed. A novel method for preparing raw point clouds for propagation simulations is presented and its positive effect on simulation accuracy demonstrated. Radio channel simulations utilizing point clouds are performed in outdoor, indoor and O2I scenarios and their accuracy validated against measurements for frequencies ranging from 4 GHz to 60 GHz. The last part of this thesis is dedicated to applications of point cloud-based radio channel simulations. Line-of-sight probability at mm-wave frequencies is evaluated for urban micro-cells in the presence of clutter typically unavailable in environment databases, e.g., pedestrians and vehicles, and a Poisson-process based model of LOS probability is verified with them. Additionally, simulated radio channel data is utilized to validate feasibility of employing coarse beam-searching performed at 4 GHz for more efficient mm-wave beam-steering at 86 GHz. Finally, site-specific simulations are utilized as a test environment for 28 GHz 5G mobile phone antenna evaluation.