Ali Chamseddine soutiendra sa thèse le jeudi 29 janvier 2026 à 15h30 en salle B014 du LISIC à Calais.
La thèse est co-dirigée par Serge Reboul, Georges Stienne et Ghaleb Faour.
Titre :
Détection de points de rupture dans la phase d’un signal : application à la télédétection par réflectométrie GNSS
Résumé :
Global Navigation Satellite System Reflectometry (GNSS-R) has emerged as a promising passive remote sensing technique, capable of providing high resolution information on
Earth’s surfaces. While most GNSS-R applications have focused on amplitude and delay observables, the carrier phase of reflected signals remains largely underexploited due to its circular nature and processing challenges. This thesis investigates the potential of carrier-phase measurements for detecting surface changes, emphasizing phase coherence variations as a sensitive indicator of environmental and geophysical dynamics. For this purpose, classical and modern change point detection (CPD) methods are reviewed and a rigorous analytical framework is developed to address the unique statistical characteristics of phase data. Circular uniformity tests are explored as indicators of coherence levels in such data. In particular, CPD based on the Rayleigh Test is applied for the detection and localization of changes in phase noise moments. A novel offline Bayesian Change Point Detection (BCPD) methodology is also adapted to circular data by modeling phase noise with the Von Mises distribution and incorporating spectral features related to excess Doppler shifts. The BCPD framework en ables principled segmentation of the data into statistically homogeneous intervals, balancing sensitivity and robustness. The proposed approaches are first validated on synthetic datasets, demonstrating, under controlled conditions, their ability to accurately detect subtle changes. Subsequently, real GNSS-R measurements are analyzed, showing that the proposed methods can successfully identify phase transitions corresponding to changes in reflecting surface types. These results illustrate the practical applicability of carrier-phase-based detection and its superiority in
capturing fine-grained surface changes. This work bridges the methodological gap between GNSS R signal processing, directional statistics, and Bayesian inference, establishing a unified framework for high-resolution, phase-based remote sensing. The findings provide both theoretical and practical contributions, paving the way for advanced environmental monitoring and geophysical applications using GNSS-R.
Jury :
M. Alexandre BAUSSARD, Université de Technologie de Troyes
M. Oussama BAZZI, Lebanese University
M. Karim ELMOKHTARI, Toronto Metropolitan University
M. Hamza ISSA, NovAtel Inc