Physically Consistent Indoor Wireless Channel Modelling and Optimisation with Ray Tracing

Polarisation is important in modelling wireless propagation, particularly indoors, where signals undergo multiple reflections and transmissions through complex building materials. Conventional ray tracing predicts propagation paths effectively, but often represents polarisation with insufficient accuracy. To address this limitation, this thesis proposes a physically consistent three-dimensional ray-tracing framework that tracks polarisation through a rotation transfer matrix. The framework follows each polarisation component through successive interactions, enabling estimation of polarisation-dependent received power and cross-polarisation ratio while preserving electromagnetic transformations.
Using this framework, the thesis investigates how indoor depolarisation affects wireless channels and polarisation shift keying. The results show that assigning a constant cross-polarisation ratio to multipath components cannot represent the variation caused by propagation conditions and may yield inaccurate average bit error-probability predictions. The framework supports an optimisation procedure that jointly adjusts the complex permittivity and thickness of multilayer materials. Within the cases studied, the optimised material configurations reduce the average bit error probability of the polarisation shift keying system at high signal-to-noise ratios and lower the resulting error floor.
The framework is then extended to reconfigurable intelligent surfaces under indoor multipath illumination. A circuit-based macromodel is incorporated into the ray-tracing formulation, while physical optics and diffraction grating theory describe multimode reradiation, finite-aperture diffraction, and parasitic scattering. The extended framework relates the electromagnetic behaviour of the surface to site-specific channel prediction. Comparisons with full-wave simulations and indoor measurements show agreement in the evaluated cases and demonstrate its ability to predict changes in received-power coverage and spatial received-power dispersion.
Overall, the framework provides a physically grounded, coherent, and computationally efficient basis for analysing and optimising polarised indoor wireless systems assisted by reconfigurable intelligent surfaces.