Positioning algorithms in 5G systems and experimental verification
Loading...
Date
2026-02-27
Authors
Μουταβελής, Φοίβος Απόλλων
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This thesis investigates positioning in 5G New Radio (NR) networks using Positioning Reference Signals (PRS) and presents an end-to-end processing pipeline that maps physical-layer observations to robust position estimates under realistic propagation conditions. Emphasis is placed on the dominant error sources in multipath propagation, non-line-of-sight (NLOS) conditions, noise, and synchronization uncertainty. An OFDM/PRS channel observation model is formulated and used to extract key channel parameters, focusing on joint Time of Arrival (TOA) and Angle of Arrival (AOA) estimation. High-resolution estimators are studied, with particular attention to MUSIC (including preprocessing for coherent multipath) and the nonparametric IAA-APES method. Their performance is evaluated through ray-tracing simulations and experimental cabled and OTA measurements, highlighting strengths in resolution and robustness. Finally, the extracted measurements are fused using conventional geometric localization methods. In parallel, an iterative LOS/NLOS positioning algorithm is developed, incorporating a hypothesis test (GLRT) to reject unreliable multipath components. Results indicate that single base-station positioning can be feasible in favorable environments, while multiple base-station setups substantially improves solvability, coverage, and overall accuracy.
Description
Keywords
5G NR positioning, Channel parameter estimation, Array signal processing, MUSIC, IAA-APES, Coherent multipath, Ray-tracing channel simulation, Experimental channel measurements