Skip to Main content Skip to Navigation

Estimation dynamique non-linéaire de canaux de transmission pour récepteurs satellites mobiles

Abstract : This thesis deals with nonlinear Bayesian estimation methods, and its application to synchronization for satellite communications systems. Moreover, we compute Bayesian bounds for the oversampled synchronization problem. First, we introduce Kalman and particle filtering, and we propose a new overview of sigma-point filters. Then, we propose a modelisation for the oversampled synchronization problem, and we compute two Bayesian bounds: the Bayesian Cramér-Rao bound for phase estiamtion and the hybrid Cramér-Rao bound for joint phase and frequency offset estimation. We apply the nonlinear filtering methods (Kalman, particle and sigma-point filters) to solve the following problems: phase estimation, joint phase and frequency offset estimation, delay estimation and phase estimation with non-Gaussian noises. The proposed methods has shown good performances for our synchronization problems. We also present in this thesis three studies related with the main problem treated overall the dissertation. The firs one concerns the joint delay and complex gains estimation for a slowly varying Rayleigh channel for CPM signals. On the second one, we use sigma-point filtering techniques for sensor data fusion for positioning applications. Finally, we present a new approach to ultra-tight GNSS/INS integration and the solution using a cubature method.
Complete list of metadatas

Cited literature [141 references]  Display  Hide  Download
Contributor : Jordi Vilà-Valls <>
Submitted on : Monday, November 8, 2010 - 11:40:47 AM
Last modification on : Thursday, November 19, 2020 - 12:59:39 PM
Long-term archiving on: : Friday, October 26, 2012 - 3:11:22 PM


  • HAL Id : tel-00533731, version 1



Jordi Vilà Valls. Estimation dynamique non-linéaire de canaux de transmission pour récepteurs satellites mobiles. Traitement du signal et de l'image [eess.SP]. Institut National Polytechnique de Grenoble - INPG, 2010. Français. ⟨tel-00533731⟩



Record views


Files downloads