Scientific direction Development of key enabling technologies
Transfer of knowledge to industry

PhD : selection by topics

Applying machine learning to improve Intrusion Detection Systems

DPACA (CTReg)

Autre

01-09-2018

SL-DRT-18-0617

pierre-alain.moellic@cea.fr

The proliferation and growing complexity of cyber-attacks targeting the networks of companies, institutions or industrial infrastructures is a major security issue. Today, it is essential to propose technological solutions to detect complex and usually new attacks more particularly for critical infrastructures such as Cyber Physical Systems (CPS) gathering strong operational constraints. Among the available security tools, intrusion detection systems (IDS) rapidly become indispensable solutions such as traditional firewalls or antivirus. However, the available solutions cannot completely thwart current threats mainly because of a detection paradigm that is focused on known attacks (misuse-based or signature-based IDS). The future of these systems go through the development of other approaches (anomaly-based IDS) and the use of analysis and modelling tools based on Machine Learning. A lot of academic works have been proposed in this sense, supported by the strong emulation in ?artificial intelligence?. However, proposed technologies suffer from a lack of real data enabling an efficient evaluation of the performances. In the highly critical context of CPS which we need to precisely define the architectures, the supervision processes and threat models, the PhD aims at developing innovative IDS solutions (using well-known open source platforms) using approaches based on Machine Learning and using real data (from CEA Cadarache). The proposed solutions will have to meet strong performance requirements (accuracy rate, false-positive/false-negative rates) to demonstrate the pertinence of these approaches for real infrastructures.

Télécharger l'offre (.zip)

Hyperspectral microscopy and single-shot optical coherence tomography with a static Fourier transform imaging spectrometer

Département Microtechnologies pour la Biologie et la Santé (LETI)

Laboratoire Imagerie et Systèmes d'Acquisition

01-09-2018

SL-DRT-18-0621

jean-charles.baritaux@cea.fr

Fourier Transform Spectroscopy measures the degree of coherence of light to recover the spectrum. A Fourier Transform spectrometer is said static when the fringe pattern is recorded in a single shot with no displacement of mechanical parts. Recently this concept was extended to Hyperspectral Imaging (HSI) for Space applications using a new configuration of static interferometer positioned in front of a focal plane array. Aside from HSI, another possibility that has not yet been investigated is to use this static interferometer for Optical Coherence Tomography (OCT). This PhD project is a collaboration between the Department of Astrophysics of the University of Grenoble and CEA Leti. We propose to investigate this new OCT approach, and its coupling to HSI in a fast bimodal system that could address many applications in Diagnostic and Bioimaging. The student will work on the development of a microscope integrating this new kind of interferometer, as well as the numerical processing of the interference patterns. Applications from students with a solid background in optics and data processing are welcome. A strong interest in biophotonics, and bioimaging is expected.

Télécharger l'offre (.zip)

Dimensioning of new X-ray multisource architectures and development of tomosynthesis reconstruction algorithm

Département Microtechnologies pour la Biologie et la Santé (LETI)

Laboratoire Détecteurs

01-09-2018

SL-DRT-18-0625

vincent.moulin@cea.fr

The imminent advent on the market of new X-ray multi-sources will allow to easily dispose of numerous additional angles of view which, via a reconstruction by tomosynthesis, will provide 3D images of the inspected objects. This modality of examination can satisfy the need of many applications for which X-ray radiography (1D projection of the object) is too restricted and tomography (3D imaging) too constraining to implement. The objective of the thesis consists, on the one hand, of imagining and simulating the first multi-sources architectures and, on the other hand, of implementing and evolving reconstruction algorithms adapted to obtain the best 3D imaging performance. The profile of the candidate for this thesis is oriented "Information processing" with however a "Physical" connotation for the understanding of phenomena of radiation-matter interaction.

Télécharger l'offre (.zip)

Flexible piezoelectric nanosystems: design, assembly and tests of an integrated sensor matrix

Département Systèmes

Laboratoire Autonomie et Intégration des Capteurs

01-09-2018

SL-DRT-18-0626

elise.saoutieff@cea.fr

The aim of the PhD thesis is to implement a matrix of flexible piezoelectric nanosensors, which enable the 3D reconstruction of a force or deformation field. The nanosensors based on GaN nanowires obtained by directed growth are fabricated and assembled at CEA. The candidate will tackle experimental aspects, which include the fabrication and the assembly of sensors and sensor networks (matrix) via controlled growth and deposition processes, first-level flexible electronic layers (interconnects), system integration on an object (mechatronics) and finally signal collection and processing through a dedicated reading electronics, to be designed based on the competences present in our laboratory. In parallel, the candidate will carry out studies at the fundamental level, such as investigating the mechanical transfer between the nanowire and its environment and its effect on the generated signal under deformation, or the study of the piezoelectric / pyroelectric coupling intrinsic to GaN nanowires. For this purpose, the candidate will have access to multi-physics simulation tools. Finally, investigations on the choice of materials and the characterisation thereof (structural, mechanical, thermal, optical, electrical) will be pertinent and may pursued. More generally, this PhD thesis will also provide the opportunity to develop applicable solutions in various fields such as deformation and impacts detectors for predictive maintenance, sensitive surfaces or electronic skin.

Télécharger l'offre (.zip)

Use of parsimony and machine learning for the real time simulation of realistic nondestructive testing signals

Département Imagerie Simulation pour le Contrôle (LIST)

Laboratoire Simulation et Modélisation en Electro-magnétisme

01-10-2018

SL-DRT-18-0628

roberto.miorelli@cea.fr

Many sophisticated numerical solvers are nowadays available for the simulation of complex configurations in the field of non-destructive testing (NDT). This complexity can come from the piece 3D geometry, for products of additive manufacturing processes, or from its composition, for composite structures or complex heat-treated steel, for instance. Beside these aspects, which motivate many efforts of theoretical modelling and numerical implementation, other environmental factors may have a strong impact on the physical NDT signals. Among these effects, one can cite other physical effects like shifts of temperature, electromagnetic perturbations coming from neighbouring machines or more simply additional physical effects that the model does not account for. These effects translate into an uncertainty on the measurements themselves (in terms of repeatability for instance) and consequently in a mismatch between simulated signals and experiments. Such discrepancy can lead to misleading conclusions and estimations. Defining in a general way the discrepancy between theoretical models and real experiments is not an easy task. This component of the signal, which can be labelled as ?noise?, depends on so many factors and practical conditions that it cannot be easily modelled or described theoretically. The proposed PhD subject proposes an approach to address the issue of characterizing this noise component and incorporating it into the simulation process. The tools implemented will focus on NDT inspections of metallic or composite structures by means of ultrasonic or electromagnetic methods. The proposed strategy relies both on advanced concepts of statistical learning theory (supervised and unsupervised learning, dictionary learning and deep learning) and on advanced numerical simulation tools (finite element, boundary element and fast semi-analytical methods) developed at CEA LIST for the simulation of NDT. First, by comparing experimental data and simulations, statistical learning algorithms will allow to determine characteristic features describing the discrepancy between them. Then, contributions based on these descriptors will complement the theoretical signals coming from the physical model in order to get a simulation that compares to experiment with better accuracy, taking into account the particular conditions of experimental acquisitions. When coupled with metamodels, acting as real time substitute of the physical model, then real-time realistic simulators can be obtained. The expected results of this ambitious thesis work are numerous. First of all, realistic models can be obtained and adapted to case-dependent industrial conditions, provided that sufficient amount of experimental data is available. On can thus expect a better performance of online diagnostics and parameters estimations. Then, such models can be used for training purposes as they can in real time generate realistic signals corresponding to various imaginary flaws affecting the material under test. Lastly, such models will enhance the estimation of confidence levels associated to NDT techniques, as they will provide a more accurate description of the real signals variability.

Télécharger l'offre (.zip)

Integrated circuit modification with focalized XRays Beam and FIB

Département Systèmes

Centre d'Evaluation de la Sécurité des Technologies de l'Information

01-10-2018

SL-DRT-18-0633

stephanie.anceau@cea.fr

The understanding and conception of electronic devices require tools to modify these devices after their fabrication. A device failure or a functional bug require to modify the component behavior in order to find the corresponding problem. This modification of the integrated circuit is classically done with a Focused Ion Beam (FIB). This equipment permits to etch materials and deposit conductive or insulator materials, which allows to modify the interconnections of the integrated circuit. This operation is called circuit edit because the device can be reconfigured after his conception. Previous experimentations on ESRF focalized beam line in Grenoble allowed to demonstrate that a focalized XRays perturbation changes the state of a single NMOS transistor and a single memory cell of SRAM-EEPROM and Flash memories block in a semi-permanent manner in the electronic device. This proof of concept has been realized on a new CMOS technology device (45 nm). The 50 nm focalization allows to modify one single NMOS transistor. The most aggressive technologies (<20nm) can be addressed with this technique even with a 50nm focalization diameter. Contrary to the FIB the interconnections of the device are not modified: the state of one (or several if necessary) transistor(s) is modified. This modification is semi-permanent because it is reversible with a simple heating treatment. This new circuit edit technique is very promising. The aim of this PhD is to explore and develop this new technique of integrated circuit modification by using an XRays focalized beam at ESRF. Among the different key points to study, there are: -Precise localization of the transistor to attack with the help of fluorescence scan and GDS layout of the circuit -Modification of single PMOS transistors -Adaptation of the technique to more aggressive technology -Possibility to work without preparation of the package device. The exploration of the perturbation possibilities with laboratory XRays beam (without synchrotron) will be studied with the help of lead and tungsten protection made thanks to the FIB. During the PhD the candidate will have access to beam line shift at ESRF.

Télécharger l'offre (.zip)

69 Results found (Page 3 of 12)
first   previous  1 - 2 - 3 - 4 - 5  next   last

Voir toutes nos offres