Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
Modèle de l'impact des multi-trajets DME/TACAN sur les futurs récepteurs GNSS pour l'aviation civile
No full textIn the upcoming years, civil aviation will adopt a new generation of Global Navigation Satellite System (GNSS) receivers, referred to as the Dual Frequency Multi-Constellation (DFMC) receivers. Unlike the legacy GNSS receivers, DFMC GNSS receivers will be capable of processing both the GNSS L1/E1 and the GNSS L5/E5a signals. Thanks to the combination of these signals in both frequency bands, the civil aviation DFMC GNSS receivers are expected to have better performances than the legacy GNSS receivers. One challenge associated with the processing of the GNSS L5/E5a signals for a civil aviation GNSS receiver is the sharing of the same Aeronautical Radionavigation Service (ARNS) frequency band with a multitude of other signals, seen as Radio Frequency Interference (RFI) signals. Among all the RFI sources, the most powerful and threatening ones are the pulsed RFI signals, such as the Distance Measuring Equipment/Tactical Air Navigation (DME/TACAN) pulses. In the civil aviation standards, the RFI impact is modelled as the degradation of the useful GNSS carrier-to-noise-density ratio, C/N_0, induced by the RFI signals at the airborne GNSS receiver correlator input. In addition, a temporal blanker is assumed to be implemented in the Radio-Frequency Front-End block of the GNSS DFMC receiver. The role of the blanker is to mitigate the pulsed received RFI signals by zeroing the samples having an instantaneous power envelope stronger than a given threshold. However, the current C/N_0 degradation model does not consider the potential presence of DME/TACAN multipath (MP). Moreover, two recent experimental studies from 2019 and 2021 have revealed the presence of strong DME/TACAN MP in I/Q sample recordings affecting the blanker in a non-negligible way. Therefore, the first contribution of this Ph.D. thesis is the update of the C/N_0 degradation model to account for the DME/TACAN MP impact generated by ground scatterers. In particular, two models, the statistical and the fixed environment models, are developed to cover different needs: the statistical model is a low complexity model providing statistical average results, defined by closed-form formulas and designed for standardization purposes; the fixed environment model is a high complexity model able to provide the C/N_0 degradation for any singular set of DME/TACAN signal conditions, which is used in this work to inspect the limitations of the statistical model. Furthermore, the application of the statistical C/N_0 degradation model, requires the precise knowledge of the MP delay and scattered power at the aircraft’s GNSS antenna port. Therefore, the second contribution of this Ph.D. thesis is the development of a wide-band, geometrically based, hybrid deterministic-stochastic Air-Ground channel model. The main technical challenge faced by the proposed AG channel model is the high number of scatterers to be considered for en-route scenarios, since millions of scatterers in the RLOS of the aircraft are potentially generating MP. To cope with this issue, the proposed AG model is based on a statistical simplification of the Physical Optics (PO) to provide a theoretical expression of the scattered power generated by the scatterers. It must be noted that the scope of the proposed AG channel goes beyond the RFI analysis and the DME/TACAN system: the proposed model could be applied to any other system composed of a ground transmitter and a sky receiver operating in the L-band. Finally, the propagation channel model and the statistical C/N_0 degradation model are jointly applied to provide the C/N_0 degradation generated by the DME/TACAN direct and MP signals at two low-altitude operational hot-spots, JALTO (Pennsylvania, US) and TIXAK (Frankfurt, Germany). It is shown that only a few illuminated scatterers generate a scattered power above the blanking threshold and that the additional C/N_0 degradation generated by presence of the DME/TACAN MP is weaker than 0.52 dB.Dans les prochaines années, l’aviation civile adoptera une nouvelle génération de récepteurs GNSS (Global Navigation Satellite System), appelés récepteurs Multi-Constellation Double Fréquence (DFMC). Contrairement aux récepteurs GNSS actuels, les récepteurs GNSS DFMC traiteront à la fois les signaux GNSS L1/E1 et L5/E5a. Cette combinaison permettra des performances accrues par rapport aux récepteurs GNSS traditionnels. Un des défis majeurs du traitement des signaux L5/E5a est le partage de la même bande de fréquence du service de radionavigation aéronautique (ARNS) avec d’autres signaux, perçus comme des interférences radiofréquence (RFI). Parmi ces sources, les signaux RFI pulsés, tels que ceux des systèmes de mesure de distance/navigation aérienne tactique (DME/TACAN), sont les plus puissants et menaçants. Les normes d’aviation civile modélisent l’impact des RFI par la dégradation du rapport signal sur densité spectrale de bruit C/N0 du GNSS, induite par les signaux RFI à l’entrée du corrélateur du récepteur GNSS embarqué. Un système de blanker temporel est supposé dans le bloc RF du récepteur GNSS DFMC pour atténuer les RFI pulsées en rendant nul les échantillons avec une puissance instantanée au-dessus d’un seuil donné. Cependant, le modèle actuel de dégradation du C/N0 ne prend pas en compte la présence possible de multipath (MP) DME/TACAN. De plus, deux études expérimentales récentes ont révélé la présence de DME/TACAN MP forts dans les enregistrements I/Q, affectant significativement le blanker. La première contribution de cette thèse consiste donc à mettre à jour le modèle de dégradation du C/N0 en intégrant l’impact des MP DME/TACAN générés par les balises au sol. Deux modèles sont développés : un modèle statistique, à faible complexité, fournissant des résultats moyens pour la standardisation, et un modèle d’environnement fixe, à haute complexité, utilisé pour analyser les limites du modèle statistique. L’application du modèle statistique de dégradation du C/N0 nécessite une connaissance précise du retard MP et de la puissance diffusée au port d’antenne GNSS de l’avion. La deuxième contribution de cette thèse est donc le développement d’un modèle de canal Air-Sol hybride large bande, basé sur une approche géométrique déterministe-stochastique. Le défi principal de ce modèle AG est le nombre élevé d'obstacles à considérer pour les scénarios en route, car des millions de diffuseurs dans la ligne de visée de l’avion peuvent potentiellement générer des MP. Pour surmonter ce défi, le modèle AG proposé repose sur une simplification statistique de l’optique physique (PO) pour fournir une expression théorique de la puissance diffusée. Il est important de noter que la portée du modèle AG dépasse l’analyse RFI et le système DME/TACAN, et pourrait être appliqué à tout autre système composé d’un émetteur au sol et d’un récepteur aérien opérant dans la bande L. Enfin, le modèle de canal de propagation et le modèle statistique de dégradation du C/N0 sont appliqués conjointement pour évaluer la dégradation du C/N0 due aux signaux directs et MP DME/TACAN à deux positions opérationnels de basse altitude, JALTO (Pennsylvanie, États-Unis) et TIXAK (Francfort, Allemagne). Il est démontré que seuls quelques obstacles illuminés génèrent une puissance supérieure au seuil du blanker, et que la dégradation supplémentaire du C/N0 due aux MP DME/TACAN est inférieure à 0,52 dB
Port-Hamiltonian formulations for the modeling, simulation and control of fluids
No full textInternational audienceThis paper presents a state of the art on port-Hamiltonian formulations for the modeling and numerical simulation of open fluid systems. This literature review, with the help of more than one hundred classifiedreferences, highlights the main features, the positioning with respect to seminal works from the literature on this topic, and the advantages provided by such a framework. A focus is given on the shallow water equations and the incompressible Navier–Stokes equations in 2D, including numerical simulation results. It is also shown how it opens very stimulating and promising research lines towards thermodynamically consistent modeling and structure-preserving numerical methods for the simulation of complex fluid systems in interaction with their environment.<br /
Évaluation et optimisation de protocoles cryptographiques post-quantiques pour les drones civils
No full textIn the context of a very significant growth in their number, it is necessary to prevent incidents and accidents in the field of UAVs. In particular, it is urgent to protect them efficiently against attacks by third parties. The on-board ground communication links in particular represent vulnerable segments. In this respect, we can consider the "command and control" messages on the C2link of the same name, but also the sensitive messages of new applications such as RemoteID, which will allow remote identification. Generally speaking, these are so-called "surveillance" messages in the context of air transport and drones. One of the main objectives of this thesis will be to integrate efficient and robust mechanisms to guarantee confidentiality, authentication, and integrity. These three primitives are mainly encryption based. In order to avoid the problem of key exchange, hybrid encryption will be considered at first. These encryptions allow the use of a "session" key, which does not need to be stored in long-term memory, thus reducing the resources used. Generally speaking, existing solutions do not seem to fully meet or will show their limits in the context of UAV systems. Thus, it is relevant and legitimate to dedicate research works aiming at designing and then evaluating the performances of cryptographic protocols for civilian UAV communications.Dans le contexte d'une croissance très importante de leur nombre, il convient de prévenir les incidents et accidents dans le domaine des drones. En particulier, il est urgent de les protéger efficacement contre les attaques dont ils pourraient faire l'objet par des tiers. Les liens de communication sol bord notamment représentent des segments vulnérables. A ce titre on peut considérer les messages de “commande et contrôle” sur le lien C2link du même nom mais aussi les messages sensibles des nouvelles applications telle par exemple que RemoteID qui permettra l'identification à distance. D'une manière plus générale il s'agit de messages dits “de surveillance” dans le contexte du transport aérien et des drones. Un des principaux objectifs de cette thèse sera d'intégrer des mécanismes efficaces et robustes afin de garantir confidentialité, authentification, et intégrité. Ces trois primitives reposent principalement sur le chiffrement. Afin de s'abstraire de la problématique d'échange de clés, dans un premier temps des chiffrements hybrides seront considérés. Ces chiffrements permettent l'utilisation d'une clé dite “de session”, qu'il n'est pas nécessaire de stocker en mémoire à long terme, réduisant ainsi les ressources utilisées. D'une manière générale, les solutions existantes ne semblent pas répondre pleinement ou montreront leurs limites dans le contexte des systèmes drones. Ainsi il est pertinent et légitime de dédier des travaux de recherche visant à concevoir puis évaluer les performances de protocoles cryptographiques pour les communications des drones civils
Constrained Zonotope-based localization of mobile agents with uncertain observation times
No full textSubmitted to IEEE TCST journal on 09/12/2024This paper presents an extended set-membership state estimator based on constrained zonotopes, addressing the practical challenge of uncertain asynchronous observation arrival times, including timing errors. Additionally, the proposed estimator accounts for parametric uncertainty in the observation equation and incorporates multiple state propagation models. The filter provides valid state enclosures despite observation time uncertainty, which is a critical feature for real-world applications. To demonstrate its effectiveness, we present an academic problem simulating a real-world application. Performance comparisons with existing nonlinear estimation algorithms, including the ellipsoidal set state estimator, Iteration Estimation Set Membership Filter (IESMF), the set inversion algorithm (SIVIA), and the classic Extended Kalman Filter (EKF), highlight the superior efficiency and performance improvements of the proposed approach
Mathématiques appliquées pour le traitement du signal: Slides pour le cours ENAC de mathématiques appliquées pour le traitement du signal destiné aux élèves de première année du cursus ISESA
No full textLicenc
A Data-Driven Framework for Modelling Complexity in Terminal Manoeuvring Area
No full textInternational audienceThis paper presents an objective, data-driven framework for quantifying air traffic complexity in the TerminalManoeuvring Area (TMA) using historical ADS-B data from Singapore TMA. The motivation for developing this frameworkstems from the limitations of traditional subjective measures, which are often influenced by individual perceptions and can varysignificantly between air traffic controllers. Subjective measures may also fail to capture real-time operational demands, especiallyin complex, high-density environments such as Singapore TMA. By focusing on operational outcomes—specifically vectoring andholding patterns—the framework provides a more accurate reflection of real-time complexity. Principal Component Analysis(PCA) and k-means clustering are employed to classify complexity levels based on trajectory features such as arc lengths,curvatures, and holding durations. The results show that total arc lengths and curvatures are significant complexity factors,with extensive vectoring contributing more to TMA complexity than holding patterns. The significance of this work lies inits data-driven and objective approach to measuring air traffic complexity, offering a more accurate reflection of real-time demands compared to traditional subjective methods. Quantitative evaluations across multiple real-world scenarios validate theframework’s effectiveness, showing that TMA complexity is more strongly associated with vectoring intensity and holding patternsthan with flight density alone. This current framework can be extended to incorporate vertical profiles of arrival and departureflights and develop predictive models with practical, actionable lookahead times for real-time air traffic management
Optimising Flight Trajectories under Thunderstorm Conditions
No full textInternational audienceGiven the operational disruptions caused by thunderstorms, this paper introduces an innovative approach to flight trajectory optimisation within the context of air traffic management, with a particular focus on minimising disruptions due to adverse weather conditions. The proposed framework integrates wake turbulence separation requirements, node conflict evaluations, and real-time operational constraints to address the complexities inherent in trajectory planning. Unlike previous studies that addressed stochastic optimal control problems for conflict resolution offline, this study presents an enhanced Selective Simulated Annealing (SSA) algorithm for online conflict mitigation. A case study conducted in the approach area of Chengdu Shuangliu International Airport (CTU), involving over 900 flights, demonstrates the applicability and effectiveness of the proposed algorithm. This result underscores the algorithm's potential as a powerful tool for trajectory optimisation, offering significant improvements in operational performance in a timely manner. Additionally, sensitivity analysis of the slot shift ranges confirms the robustness of the selected parameters, highlighting their role in minimising delays and maintaining system efficiency. By deactivating decisions individually, this study also discusses the sensitivity of the algorithm to the proposed decision combinations.</div
Robotics, Artificial Intelligence and Humans: A Roadmap, or A Cheat Sheet (or Both?)
No full textInternational audienceArtificial Intelligence and Robotics are trending domains these days.Week after week, there are new impressive videos of robots who are walking, running or flipping, and AI agents performing complex tasks. Such achievements, in addition to all that is happening in related areas, could make us think that the introduction of robots into our daily lives may happen, if not tomorrow, in the near future. But from our robotics research laboratory experience, this future is not for tomorrow. But who actually knows? In any case, regardless of what will happen, it is important to prepare and be prepared for it, it is important to pursue interdisciplinary fundamental research, it is important to place the Human at the heart of our research concerns. In this paper, we propose a Fundamental and Holistic approach of Artificial Intelligence and Robotics FOR, AMONG, WITH and BY Humans
Accelerating Swarms : Harnessing Hardware Acceleration and Parallelization in Multi-Agent Reinforcement Learning
No full textInternational audienceIn Multi-Agent Reinforcement Learning, researchers often face a challenging trade-off: the use of complex environments that demand substantial computational resources, or simpler dynamics for expedited execution, albeit at the cost of transferability to more realistic tasks. This article delves into the potential of vectorizable environments, which enable parallel environment rollout and fully harness the parallelization capabilities of modern GPUs. We present a comparison of various RL environment libraries, highlighting their features and limitations for end-to-end hardware-accelerated training pipelines. We observe that the most commonly used RL algorithm libraries have yet to fully embrace end-to-end hardware-accelerated training pipeline and the limited cross-compatibility between the frameworks used for hardware acceleration and parallelization in machine learning: PyTorch, TensorFlow, and JAX, limits the mix-and-match options for RL environments and algorithms libraries
Control of the DarkO Tail-Sitter Drone through an LMI-Based Static Output Feedback Design
No full textInternational audienceThis paper deals with the control of a convertible drone through the synthesis of Static Output Feedback (SOF) model-based controllers. To that end, a convex optimization algorithm based on Lyapunov's stability theory and Linear Matrix Inequalities (LMI) available in the literature is employed.To evaluate the algorithm's performance, given that it has never been previously tested on realistic dynamical systems, it is implemented on the experimental model of the DarkO drone. Then, simulations of the closed-loop dynamic of the drone are carried out to assess the performance of the new design. For a preliminary experimental validation, the control architecture is implemented on the real drone system and test flights focused on the hovering phase and reference tracking in terms of position are carried out