Validation of Sionna Ray-Tracer for Urban Signal-Strength Prediction Using Large-Scale Drive-Test Measurements

Abstract

Accurate channel modelling is a key prerequisite for planning and improvement of wireless communication networks. Obtaining a precise channel model is therefore a pivotal task for telecommunication engineers. One method that has gained in popularity in recent years is ray-tracing. Given the widespread use, it is important to understand the strengths and weaknesses of ray-tracing and to know what performance is to be expected in praxis. In this paper, we compare a large dataset of outdoor measurements of signal strength with simulations and assess the obtained error changes given different antenna models, 3D models with different levels of detail and various simulation settings. The benefits of material calibration are also examined. Diffuse reflection and accurate digital-twin are identified as essential for signal strength prediction, whereas diffraction has a negligible impact in the considered scenario. The material calibration turned out to be beneficial, yielding a consistent mean absolute error improvement of approximately 1 dB.