André Bezerra de Freitas Diniz, Kommunikation, Antenner och Optiska Nätverk

The evolution of communication systems towards 6G envisions ubiquitous global connectivity, where non-terrestrial networks will be crucial. Low Earth orbit (LEO) satellites are particularly attractive due to favorable propagation characteristics and low latency. However, their high speeds induce severe Doppler shifts, posing synchronization challenges in satellite-to-ground communications. In addition, vulnerabilities in global navigation satellite systems highlight the need for robust localization alternatives, which are fundamental to enable a wide range of use cases. To avoid high complexity and strict synchronization demands of multi-satellite solutions, Doppler-based single-satellite methods offer an attractive, low-requirement alternative. This thesis addresses these challenges by developing novel, standalone methodologies for Doppler shift tracking and Doppler-based user localization in single LEO satellite systems.

We initially address the Doppler shift tracking problem by a multi-step Doppler shift estimation algorithm based on linear estimators. Taking advantage of additional knowledge of Kepler's laws describing satellite motion, a model-based post-processing stage is proposed. Overall, this procedure provides a continuous and smooth curve describing the Doppler shifts over the satellite's pass in degrading channel conditions.

Following, we develop a Doppler-based single-satellite localization algorithm where several impairments are jointly modeled. These are originated due to the receiver's hardware, atmospheric phenomena and imprecise knowledge of the satellite states. Higher accuracy in relation to previously developed methods in the literature, which do not address the combined effect of such perturbations, is shown. As such, our proposed technique may be applied in practical and resource-constrained scenarios.