Millimetergolf RCS-gebaseerde kanaalmodellering, CSI-feedbacksysteem, en fased arraymodule met tweevoudige bundel

This dissertation presents a novel approach to 3D millimeter-wave statistical channel modeling utilizing Radar Cross Section (RCS) as a key parameter. We introduce a quasi-deterministic ray tracing (QD-RT) method to accurately model the bistatic RCS of irregular objects, employing probability density functions for enhanced fidelity. Our method achieves comparable channel characterizations to deterministic methods, demonstrating its effectiveness in capturing complex scattering phenomena. The proposed QD-RT method can generate the same statistical channel parameters—such as path loss, root mean squared (RMS) delay spread, and angular spread—as deterministic ray tracing (D-RT), while benefiting from reduced computational complexity. We also develop a high-frequency scattering model for irregular objects such as the human body, cars, and windows using Physical Optics (PO) and Geometrical Optics (GO). Scattering from windows is calculated using iterative PO and double-bounced GO. To accelerate bistatic RCS calculations, we employ neural networks, which significantly reduce computational time while maintaining accuracy. Experimental validation is provided through a millimeter-wave channel measurement campaign in a courtyard scenario using a massive MIMO channel sounder. The analysis focuses on scattering effects from building facades, highlighting the impact of urban environments on millimeter-wave propagation characteristics. In addition, the statistical scattering model of the human body is verified at 28 GHz using the same channel sounder setup. This dissertation further introduces a novel limited channel state information (CSI) feedback system that estimates maximum received power using a joint neural network. Moreover, we propose and fabricate a low-RF hardware complexity dual-beam phased array architecture for massive MIMO applications. The proposed design reduces the number of active phase shifters by 50% while maintaining the antenna aperture, thereby improving energy efficiency and scalability. Through these innovations and empirical investigations, this dissertation advances the understanding and modeling of millimeter-wave channels, paving the way for enhanced performance and efficiency in future wireless communication systems.