Extended Kalman Filtering for DORIS : Modelling Choices and Performance Impacts
Xanthos Papanikolaou (National Technical University of Athens, Greece)
Vissarion Fisikopoulos (National Technical University of Athens, Greece); Dimitris Anastasiou (National Technical University of Athens, Greece); Maria Tsakiri (National Technical University of Athens, Greece)
Event: 2026 IDS Workshop
Session: Advances in DORIS: Data, Antennas, and Modeling
Presentation type: Poster
DOI: 10.24400/312072/i03-2026.4616
Licence: 
Abstract
Over the last few years, DSO has been developing its own in-house DORIS data processing software tools, with the aim of delivering state-of-the-art, free and open-source software to the DORIS community through an open and collaborative development model. The software architecture is built around the principles of efficiency, modularity, and reusability, supporting both precise orbit determination and positioning applications. A key distinction from software traditionally used by DORIS Analysis Centers is the adoption of an extended Kalman filter as the main framework for estimating the parameters of interest.
The present contribution focuses on the implementation of the filter and the overall estimation strategy adopted by the software, with application to DORIS data from recent missions over selected time periods. For precise applications, several aspects require careful treatment, including filter tuning, process noise definition, the choice of estimated parameters and their stochastic properties, and the modeling of the observation equations. These elements strongly influence the stability and performance of the estimation process, and their proper configuration is essential for achieving reliable results. Particular attention is given to the balance between model complexity and numerical robustness, especially in the context of operational processing.A set of representative configurations is analysed, including different process noise models and parameterizations, and their impact is evaluated through orbit metrics. Particular attention is given to the balance between model complexity and numerical robustness, especially in the context of operational processing. Α range of modeling choices and their implications for performance, robustness, and estimation quality is discussed , highlighting the practical considerations that arise in the development of a modern DORIS processing system.
Contribution: IDS26_s4_Papanikolaou-ExtendedKalmanFilteringDORISModelling(poster).pdf (pdf, 652 ko)
Corresponding author:
Xanthos Papanikolaou
National Technical University of Athens
Greece
