Localization via 6G LEO Satellites and RISs

Abstract

Modern autonomous applications, such as intelligent transportation, unmanned aerial vehicles, and remote sensing systems, require accurate and uninterrupted positioning ervices. These requirements are no longer met via conventional satellite localization systems, such as the global navigation satellite system (GNSS), as they suffer from weak received signals which can be easily jammed and/or spoofed, limited geometric diversity, and severe blockage in complex propagation environments. Compared with GNSS, 6G low-earth-orbit (LEO) satellites can provide lower propagation delay, stronger received signals, and richer Doppler information due to the LEO orbit. However, LEO-satellite-based localization still faces significant challenges, particularly in single-LEO scenarios where the line-of-sight (LoS) signal may be weak, obstructed, or insufficient to provide reliable localization geometry. 6G’s emerging reconfigurable intelligent surfaces (RISs) provide a promising solution by creating controllable reflected paths that can act as virtual anchors for localization. Although prior studies have investigated single-RIS-assisted LEO localization, the RIS-assisted path is usually weak and sensitive to noise and interference, which may cause severe performance degradation in low signal-to-noise-ratio (SNR) scenarios. To improve localization robustness, this thesis investigates a multi-RIS-assisted single-LEO localization framework. Multiple RISs are exploited to provide additional geometric constraints, while a joint positioning method is adopted to enhance positioning robustness. Moreover, a time-orthogonal RIS coding scheme is applied to separate different RIS branches and mitigate inter-RIS interference during channel parameter estimation. Simulation experiments are conducted to evaluate the localization performance of the proposed framework by comparing the theoretical Cramer–Rao bound (CRB) with the corresponding estimation results. The results demonstrate that double-RIS setup drops the 3D position error from 104 m level to 10 m level compared to single-RIS setup in low-SNR region, which provides a useful basis for future LEO-NTN localization systems.