Development of High-Performance Wireless Channel Models for Transportation Systems

One promising technology in rail signaling is communication-based train control (CBTC), offering substantial enhancements to the efficiency of light rail, subway, and high-speed train networks. In CBTC systems, telecommunications are employed between the train and track equipment for traffic management and infrastructure control. Due to the safety-critical nature of rail signaling and the rapid, cost-effective nature of system deployment, the existence of an accurate and efficient wave propagation model for transportation systems is critical. This thesis presents a high-performance wireless channel model, based on parabolic wave equation (PWE) methods, for radio wave propagation predictions in realistic and complex railway environments. The developed 3-D PWE-based tools are validated in a variety of scenarios, including waveguides, irregular terrain environments, and realistic arched tunnels. Then, a systematic analysis of the numerical error and computational complexity of various PWE methods are provided, and concrete guidelines for the choice of parameters are also given. Furthermore, the thesis also presents an efficient model for propagation prediction in tunnels by combining the PWE method with sparse Fourier transform techniques. Moreover, two machine learning (ML)-driven PWE methods are introduced, aimed at offering promising solutions to reconcile the traditional trade-off between accuracy and efficiency. Finally, a physics-based trajectory design approach for cellular-connected unmanned aerial vehicles (UAVs) in rainy environments, and an efficient localization framework in road tunnels are presented. It is shown that such modeling tools can be quite useful as they enable the design and optimization of wireless communication systems through integrating propagation modeling with network-level systems, instead of carrying out these tasks conventionally.