Portail HAL Ensta
Not a member yet
11080 research outputs found
Sort by
Conception d’un récepteur AIS détectant les falsifications de messages : développement de stratégies et prototypage sur FPGA
No full textAIS is a widely used communication system for ships. It allows the automatic exchange of information such as position, speed, course, identity, port of departure and arrival, etc. This information assists navigation, secures maritime traffic and simplifies surveillance by the coastguard. However, having been developed in the 90s, AIS is not very secure. [...]. Based on this observation, in this manuscript, three strategies have been developed to detect these attacks. The first one tracks ships with a filter to detect position and speed falsifications. The second verifies that the boats respect the TDMA access mode when they communicate. The verification of this access mode, specified by the AIS standard, makes it possible to detect the transmission of false messages and the creation of ghost ships. Finally, the last strategy calculates a radiometric signature from the signals transmitted by each transponder in order to identify them physically and not by the identity they transmit in their messages.[...]. These three strategies are applied jointly and have been tested on real signals recorded in the Brest harbour. First and second kind error rates of 0.0035 and 0.0171 respectively on the detection of identity spoofing have been obtained. In addition, Monte Carlo simulations have shown a sensitivity to position falsifications ranging from 35m to 250m. Some of these strategies have been implemented on a FPGA to be used by the coast uard or the navy. The entire AIS signal demodulation chain was also implemented on the same board to recover the signals and information used by these strategies. The generation of the code implementing the FPGA was done using the HLS which is a high level synthesis tool. This application case showed that this tool is now a mature technology that can efficiently and automatically synthesise large complete systems from high-level modelling, thus removing the separation that once existed between signal processing and hardware design activities. The resulting AIS receiver can detect position falsifications from baseband signals in real time. Finally, an FPGA-based virtual laboratory that simulates the receiver and its environment at the same time has also been developed. It allows to easily generate various test scenarios to verify the behaviour of the receiver in its limit conditions of use. The joint consideration of the receiver, its environment and their interactions led us to talk about cyber-physical systems and the difficulties, still current, associated with the modelling and simulation of such systems. A new method has been proposed to solve these difficulties. This method models the CPS as a set of concurrently executing actors, uses the FPGA as a simulation platform and applies HLS to automatically generate the hardware description from the actor based modelling. A virtual laboratory has been created for our receiver.It allows to simulate, at the same time, our receiver and its maritime environment for which the number of communicating vessels, their trajectories, the SNR of their transmitted signals and their transponder’s CFO can be modified.L’AIS est un système de communication pour bateaux très répandu. Il permet l’échange automatique d’informations de navigation comme la position, la vitesse, la route, l’identité, le port de départ et d’arrivée, etc. Ces informations aident à la navigation, sécurisent le trafic maritime et simplifient la surveillance des autorités de contrôle. Cependant, ayant été développé dans les années 90, l’AIS est peu sécurisé. [...]. Partant de ce constat, dans ce manuscrit, trois stratégies ont été développées pour détecter ces attaques. La première suit les bateaux avec un filtre IMM pour détecter les falsifications de position et de vitesse. La deuxième vérifie le respect, par les bateaux, du mode d’accès TDMA lorsqu’ils communiquent. La vérification de ce mode d’accès, spécifié par la norme AIS, permet de détecter l’émission de faux messages et la création de bateaux fantômes. Enfin, la dernière stratégie calcule une signature radiométrique, à partir des signaux transmis par chaque transpondeur, afin de les identifier matériellement et non par l’identité qu’ils transmettent dans leurs messages [...]. Ces trois stratégies sont appliquées conjointement et ont été testées avec des signaux réels enregistrés dans la rade de Brest. Des taux d’erreur de première et de deuxième espèce respectivement de 0.0035 et 0.0171 sur la détection de l’usurpation d’identité ont été obtenus. Par ailleurs, des simulations de Monte Carlo ont montré une sensibilité aux falsifications de position allant de 35m à 250m. Une partie de ces stratégies a été implémentée sur un système embarqué de type FPGA pour être utilisée par la douane ou la marine française. Toute la chaîne de démodulation des signaux AIS a aussi été implémentée sur la même carte pour récupérer les signaux et les informations des messages utilisés par ces stratégies. La génération du code implémentant le FPGA s’est faite en utilisant la HLS qui est un outil de synthèse haut niveau. Ce cas d’application a montré que cet outil était aujourd’hui une technologie mature pouvant synthétiser efficacement et automatiquement des systèmes complets de grande taille à partir d’une modélisation haut niveau, supprimant ainsi la séparation qui existait autrefois entre les activités de traitement de signal et les activités de conception matérielle. Le récepteur AIS intelligent ainsi conçu peut, à partir de signaux en bande de base, détecter, en temps réel, les falsifications de position. Pour finir, un laboratoire virtuel sur FPGA, simulant, en même temps, le récepteur et son environnement, a été aussi développé. Il permet de générer facilement des scénarios de test variés afin de vérifier le comportement du récepteur dans ses conditions limites d’utilisation. La considération conjointe du récepteur, de son environnement et de leurs interactions nous a amené à parler des systèmes cyber-physiques et des difficultés, toujours actuelles, associées à la modélisation et la simulation de tels systèmes. Une nouvelle méthode a été proposée pour résoudre ces difficultés. Cette méthode modélise le CPS comme un ensemble d’acteurs s’exécutant de manière concurrente, utilise le FPGA comme plateforme de simulation et applique la HLS pour générer automatiquement la description matérielle à partir de la modélisation à base d’acteurs. Un laboratoire virtuel a été créé sur FPGA pour vérifier le comportement de notre récepteur. Il permet de simuler en même temps notre récepteur et son environnement maritime, pour lequel le nombre de bateaux communiquant, leur trajectoire, le SNR de leurs signaux transmis et le CFO de leur transpondeur peuvent être modifiés. En plus, des falsifications de position et des usurpations d’identité peuvent aussi être ajoutées aux messages échangés durant la simulation
On the ability of a viscoelastic model to predict the average cyclic energy dissipated by a 3D layer-to-layer composite
No full textInternational audienceThis article aims at providing a better understanding of the dissipative mechanisms activated in a 3D layer-to-layer composite under relatively low cyclic loadings. This is done by comparing the numerical and experimental dissipation responses of the investigated specimen under the so called heat build-up protocol. For this purpose, a finite element model based on micro-computed tomography results of the tested specimen was generated. The simulation/experiment comparisons show that the proposed generalised Maxwell model predicts well the average dissipation as long as the specimen is free of damage. In this regard, the first experimentally detected thermal events enable to highlight clearly the limits of the adopted modelling strategy
High average gradient in a laser-gated multistage plasma wakefield accelerator
No full textInternational audiencePlasma wakefield accelerators driven by particle beams are capable of providing accelerating gradient several orders of magnitude higher than currently used radio-frequency technology, which could reduce the length of particle accelerators, with drastic influence on the development of future colliders at TeV energies and the minimization of x-ray free-electron lasers. Since inter-plasma components and distances are among the biggest contributors to the total accelerator length, the design of staged plasma accelerators is one of the most important outstanding questions in order to render this technology instrumental. Here, we present a novel concept to optimize inter-plasma distances in a staged beam-driven plasma accelerator by drive-beam coupling in the temporal domain and gating the accelerator via a femtosecond ionization laser
Preparing engineering students for the challenges of the SDGs: what competences are required?
No full textInternational audienceDespite the emerging discussions about the growing role of engineers in achieving the Sustainable Development Goals (SDGs), there is a lack of agreement on which competences should be prioritised to prepare engineering students to resolve future sustainability challenges. This study examined and compared the views of key stakeholders of engineering education (Academics, Employers and Students) using twelve focus groups in Denmark, Finland, France and Ireland. The findings were mapped against competences identified in previous studies to highlight gaps and opportunities for development. The results confirm the strong emphasis on normative, strategic and systems thinking competences in engineering. However, the outcomes also lack acknowledgement of anticipatory competence, contradicting the future oriented perspective required to achieve sustainable development. The findings can be used by educators to inform programme development and to implement opportunities for students to develop the competences necessary to support sustainable development and the SDGs
A Quantum Algorithm for the Sub-Graph Isomorphism Problem
No full text28 pages, 11 figuresInternational audienceWe propose a novel variational method for solving the sub-graph isomorphism problem on a gate-based quantum computer. The method relies (1) on a new representation of the adjacency matrices of the underlying graphs, which requires a number of qubits that scales logarithmically with the number of vertices of the graphs; and (2) on a new Ansatz that can efficiently probe the permutation space. Simulations are then presented to showcase the approach on graphs up to 16 vertices, whereas, given the logarithmic scaling, the approach could be applied to realistic sub-graph isomorphism problem instances in the medium term
A Lightweight Method to Define Solver-Agnostic Semantics of Domain Specific Languages for Software Product Line Variability Models
No full textInternational audienceWe propose a method to address the current lack of standards for both software product line variability modeling languages and their formal semantics. It allows specifying, in an agile, declarative, and solver-agnostic fashion the formal semantics of a domain-specific variability modeling language through a simple JSON based specification format. Our approach leverages the Common Logic Interchange Format (CLIF) standard for interoperability among logical inference engines. We demonstrate our approach with two concrete examples of Variability Models, and present the tooling and architecture that makes this possible
Modélisation des neurones, équation de Bloch-Torrey et leur application à l'estimation de la microstructure du cerveau par IRM de diffusion
No full textNon-invasively estimating brain microstructure that consists of a very large number of neurites, somas, and glial cells is essential for future neuroimaging. Diffusion MRI (dMRI) is a promising technique to probe brain microstructural properties below the spatial resolution of MRI scanners. Due to the structural complexity of brain tissue and the intricate diffusion MRI mechanism, in vivo microstructure estimation is challenging.Existing methods typically use simplified geometries, particularly spheres, and sticks, to model neuronal structures and to obtain analytical expressions of intracellular signals. The validity of the assumptions made by these methods remains undetermined. This thesis aims to facilitate simulationdriven brain microstructure estimation by replacing simplified geometries with realistic neuron geometry models and the analytical intracellular signal expressions with diffusion MRI simulations. Combined with accurate neuron geometry models, numerical dMRI simulations can give accurate intracellular signals and seamlessly incorporate effects arising from, for instance, neurite undulation or water exchange between soma and neurites.Despite these advantages, dMRI simulations have not been widely adopted due to the difficulties in constructing realistic numerical phantoms, the high computational cost of dMRI simulations, and the difficulty in approximating the implicit mappings between dMRI signals and microstructure properties. This thesis addresses the above problems by making four contributions. First, we develop a high-performance opensource neuron mesh generator and make publicly available over a thousand realistic cellular meshes.The neuron mesh generator, swc2mesh, can automatically and robustly convert valuable neuron tracing data into realistic neuron meshes. We have carefully designed the generator to maintain a good balance between mesh quality and size. A neuron mesh database, NeuronSet, which contains 1213 simulation-ready cell meshes and their neuroanatomical measurements, was built using the mesh generator. These meshes served as the basis for further research. Second, we increased the computational efficiency of the numerical matrix formalism method by accelerating the eigendecomposition algorithm and exploiting GPU computing. The speed was increased tenfold. With similar accuracy, the optimized numerical matrix formalism is 20 times faster than the FEM method and 65 times faster than a GPU-based Monte Carlo method. By performing simulations on realistic neuron meshes, we investigated the effect of water exchange between somas and neurites, and the relationship between soma size and signals. We then implemented a new simulation method that provides a Fourier-like representation of the dMRI signals. This method was derived theoretically and implemented numerically. We validated the convergence of the method and showed that the error behavior is consistent with our error analysis. Finally, we propose a simulation-driven supervised learning framework to estimate brain microstructure using diffusion MRI. By exploiting the powerful modeling and computational capabilities that are mentioned above, we have built a synthetic database containing the dMRI signals and microstructure parameters of 1.4 million artificial brain voxels. We have shown that this database can help approximate the underlying mappings of the dMRI signals to volume and surface fractions using artificial neural networks.L'estimation non invasive de la microstructure du cerveau, qui se compose de nombreux neurites, de somas et de cellules gliales, est essentielle pour l'imagerie cérébrale. L'IRM de diffusion (IRMd) est une technique prometteuse pour sonder les propriétés microstructurelles du cerveau en dessous de la résolution spatiale des scanners IRM. En raison de la complexité structurelle du tissu cérébral et du mécanisme complexe de l'IRM de diffusion, l'estimation de la microstructure in vivo est un défi. Les méthodes existantes utilisent généralement des géométries simplifiées, notamment des sphères et des bâtons, pour modéliser les structures neuronales et obtenir des expressions analytiques des signaux intracellulaires. La validité des hypothèses faites par ces méthodes reste indéterminée. Cette thèse vise à faciliter l'estimation de la microstructure du cerveau par simulation en remplaçant les géométries simplifiées par des modèles réalistes de la géométrie des neurones et les expressions analytiques des signaux intracellulaires par des simulations d'IRM de diffusion. Combinées à des modèles précis de la géométrie des neurones, les simulations numériques d'IRMd peuvent donner des signaux intracellulaires précis et incorporer les effets dus, par exemple, à l'ondulation des neurites ou à l'échange d'eau entre le soma et les neurites.Malgré ces avantages, les simulations d'IRMd n'ont pas été largement adoptées en raison de l'inaccessibilité des fantômes numériques, de la faible efficacité de calcul des simulateurs d'IRMd et de la difficulté d'approximer les mappings implicites entre les signaux d'IRMd et les propriétés de la microstructure. Cette thèse contribue à la résolution des problèmes susmentionnés de la manière suivante : (1) en développant un générateur de maillage de neurones open-source et en rendant accessibles au public plus d'un millier de maillages cellulaires réalistes ; (2) en augmentant d'un facteur dix l'efficacité de calcul de la méthode du formalisme matriciel numérique ; (3) en mettant en œuvre une nouvelle méthode de simulation qui fournit une représentation de type Fourier des signaux IRMd ; (4) en proposant un cadre d'apprentissage supervisé basé sur la simulation pour estimer la microstructure du cerveau par IRM de diffusion
Lecture notes on numerical linear algebra
No full textEngineering schoolCourse notes, ENSTA Paris (2nd year and M1 level), 2021
Pour dégenrer le numérique, cassons les codes !
No full textInternational audienceLe numérique a attrapé un sexe » et il est masculin. A l’heure où ce champ professionnel a pris une place cruciale dans notre économie, la faible proportion de femmes en son sein va peser durablement sur l’égalité professionnelle entre les femmes et les hommes. Bien plus, le numérique constitue aujourd’hui un champ de pouvoir et d’innovation majeure et les femmes y sont sous-représentées. Après avoir cerné les éléments explicatifs à cet état de fait, nous proposerons d’identifier les solutions possibles en centrant notre propos sur l’intérêt des rôles modèles qui agissent comme figure d’identification pour les jeunes femmes à des moments clés de leur orientation scolaire
A non-standard Halo Mass Function as a solution to the structure-growth tension, application to KiDS-1000 and DES-y3
No full textInternational audienceSemi-analytical computation of the matter power spectrum often relies on the halo mass function (HMF) as a key component. In this paper, we explore how certain variations of the HMF affect the modelling of the matter power spectra and quantify the impact on the structure growth parameter . We use the weak gravitational lensing 2-point correlation functions from both the KiDS-1000 and DES-y3 to constrain the HMF parameters, which are sensitive to dark matter properties, structure formation, and baryonic feedback. We find that, when assuming a Planck background cosmology, the canonical HMF parameters are ruled out at more than for both KiDS and DES. In the reconstructed HMF from these posteriors, we observe a lower for KiDS-1000 and a lower for DES-y3 in terms of total halo mass larger than , when compared to a canonical HMF model. In addition, a higher, Planck-like is also preferred if massive haloes were to have a more than lower abundance compared to a canonical halo mass function. Under one of these alternative HMFs, we found a rescaled for KiDS-1000 and for DES-y3. Our work suggests that varying the halo abundance provides an alternative to varying the matter power spectrum when exploring possible solutions to the tension. This difference in halo abundance may come from both dark matter particle physics or astrophysical processes, such as a lower halo merging rate or a stronger active galactic nuclei feedback. By comparing the posteriors of both cosmological and HMF parameters between two different surveys, we are also testing the self-consistency of the cosmological interpretation at a level that has corresponding particle and/or astrophysical interpretations