HAL Portal IOGS (nstitut d'Optique Graduate School)
Not a member yet
12589 research outputs found
Sort by
Federated Representation Learning for Indoor-Outdoor Detection in beyond 5G networks
No full textInternational audienceScarcity of labelled datasets makes it challenging to train robust mobile user environment detection models. Labelling and centralizing large data amounts for training is expensive. To address these issues, semi-supervised learning techniques aim to reduce data labelling, while Federated Learning (FL) avoids centralizing the data. In this work, we propose a novel approach for Indoor/Outdoor detection by combining the strengths of federated and semi-supervised learning. It consists of 2 steps: (1) Unsupervised Federated representation Learning to learn representations using large unlabelled data. We leverage unlabelled data from diverse sources situated across various geographical locations. Through FL, we develop high-quality representations by jointly learning from this distributed unlabelled data. (2) We then capitalize on the acquired representations and further employ transfer learning to achieve accurate detection using a reduced amount of labelled data. We also add an optimization module referred as User Behavioural Optimizer that corrects environment detection errors by tracking behavioural anomalies. We obtain an F1-score of 95.06% using only 30% of the entire amount of labelled data available
Ultrafast laser-induced surface complexity patterns at the nanoscale
No full textInternational audienceUltrashort laser sources (100 fs pulse duration) concentrate a large number of photons in time and space, enabling them to transform and sculpt any solid surface achieving ultimate scales of structuring down to 100 nm [1]. Upon multi-shot irradiation the self-organization and growth of periodic patterns arise from small, localized perturbations of the underlying optical coupling on random surface nanoreliefs. Brought far from equilibrium under successive ultrafast laser pulse photoexcitation, the material progressively exhibits periodic pattern structure that reveal signatures of complexity. Some laser-induced surface structures have periods typic from plasmonic and nonlinear optics origin while others show symmetry breaking characteristics of fluid dynamics, interrogating on the main mechanism that drives the organization nature. In particular, a flat surface turns into a forest of nanopeaks with a remarkably high aspect-ratio (5:1) and a sub-100 nm periodicity. Finally, the one-step approach to fabricate high density of nanocavities or nanopeaks offers a promising way to design and engineer surface properties with nanofeatures surpassing those of any naturally-occurring surfaces with expected innovative applications in biomedecine, metaphotonics and nanocatalysis [2].From the fundamental aspect, one of the challenges is to develop a general model that inherit relevant symmetry and scale invariance properties and that contain the nonlinear dynamics able to reproduce dissipative structures in spatially extended systems. A nonlinear dynamics modelling is proposed to reproduce hydrodynamic fluctuations at the onset of convective instability that we have recently demonstrated as the very nature of the laser-induced self-organized nanopatterns. I will show that the complexity of surface 2D patterns emergence can be finally learned by a deep convolutional network to connect the model coefficients to the experimental irradiation conditions, providing key laser process parameters to design a specific pattern [3].[1] R. Stoian and J.P. Colombier, “Advances in ultrafast laser structuring of materials at the nanoscale”, Nanophotonics 9(16), 4665-4688 (2020). [2] A. Nakhoul, A. Rudenko, C. Maurice, S. Reynaud, F. Garrelie, F. Pigeon, and J.P. Colombier, “Boosted Spontaneous Formation of High-Aspect Ratio Nanopeaks on Ultrafast Laser-Irradiated Ni Surface”, Advanced Science, 9 (21) 2200761 (2022).[3] E. Brandao, A. Nakhoul S. Duffner, R. Emonet, F. Garrelie, A. Habrard, F. Jacquenet, F. Pigeon, M. Sebban. & J.P. Colombier “Learning complexity to guide light-induced self-organized nanopatterns”, Physical Review Letters, in press (2023)
Microarchitectural Insights into Unexplained Behaviors under Clock Glitch Fault Injection
No full textInternational audienceWith the widespread use of embedded system devices, hardware designers and software developers started paying more attention to security issues in order to protect these devices from potential threats. Physical attacks represent an important threat to these devices, and fault injection is one of the major physical attacks. However, misunderstanding the effects of the fault injection would lead to proposing either over-protections or under-protections for these devices, thus affecting the performance/cost ratio and/or the security of the device. In this article, we provide a better representation of occurring fault, as a result of clock glitch, through novel models, in order to better understand the effects of fault injection. Also, we examine their dependencies with respect to the target device and the target program. Finally, we make use of the presented fault models to break the control-flow integrity of a program by altering the value of the program counter, in order to provide an actual application example
Multi-Scale Simulation of Oxidation Effects on Wetting Properties of Femtosecond Laser-treated Surface
No full textInternational audienceUltra-short laser pulses are used to texture various surfaces producing micro- and nano-scale reliefs on materials resulting in an increased surface area and an enhanced surface energy [1]. This kind of surface treatment has many advantages, such as the reduction of thermal effects and the minimization of surface contamination. These textured surfaces have found diverse applications in viruses and bacteria repulsion, including the osseointegration process in biomedical implants [2]. One of the known challenges is also the fact that surface properties can be unstable and evolve with time. In fact, despite the high ver-satility and efficiency of laser treatment, it can favor surface oxidation or induce chemical modifications leading to changes in wettability over time. The underlying reasons for these changes are, however, not yet understood [3]. To shed light to these effets, we have performes a series of reactive all-atom molecular dynam-ics (MD) simulations to investigate the impact of early oxidation on laser-treated titanium (Ti). Additionnaly, computational fluid dynamics (CFD) is used to quantify droplet behavior on laser-textured metal surfaces [4]. The MD simu-lations were carried out at several temperatures ranging from 100K to 1000K. The obtained calculation results show how he thickness of the resulting oxide lyers increases with time at different temperatures (fig. 1a). Among the results, we note that the TiOx layer exhibits faster growth in the beggining (at 0–130 ps), as compared to the later stages (130–300 ps). The obtained results are then employed in a series of wetting simulations, including droplet behavior on Ti and Ti with TiOx lay-er/inclusions. The simulations reveal that the presence of oxide reduces the hydrophobicity of the droplet, which is consistent with several experimental observations. In fact, the contact angle decreases over time (Fig 1b.) due to the increasing thickness of the oxide layer [4]. This effect becomes more pronounced at higher temperatures [5]. In summary, the analysis of the obtained resultsprovide several new explanations of surface wettability switching and time-evolution after ultra-short laser treatments. The dynamics of the early titanium oxidation, as determined in our simulations, correspond to the early time-evolution of surface wettability in laser-treated materials.Keywords: femtosecond laser, oxidation, wettability, reactive molecular dynamics, computational fluid dynamics.(a) (b)Figure 1: (a) Oxidation process of Ti, (b) Wet-tability of Ti and TiOx.References:[1] Cunha, A., Anne-Marie E., Laurent P., Ana P. S., Ana M. B. Rego, Amélia A., Maria C., Marie-Christine D., and Rui V. Applied Surface Science 360 (2016): 485-493.[2] Samanta, A., Wang, Q., Shaw, S. K., & Ding, H. (2020). Materials & Design, 192, 108744.[3] Omeje, I. S., & Itina, T. E. (2022). Applied Surface Science Advances, 9, 100250.[4] Lavisse, L., Grevey, D., Langlade, C., & Vannes, B. Applied Surface Science 186.1-4 (2002): 150-155.[5] Vyas, Vandan V., and Kamlesh V. Chauhan. Materials Performance and Characterization 9.1 (2020): 638-645
Beyond the disk: EUV coronagraphic observations of the Extreme Ultraviolet Imager on board Solar Orbiter
No full textInternational audienceContext. Most observations of the solar corona beyond 2 R ⊙ consist of broadband visible light imagery carried out with coronagraphs. The associated diagnostics mainly consist of kinematics and derivations of the electron number density. While the measurement of the properties of emission lines can provide crucial additional diagnostics of the coronal plasma (temperatures, velocities, abundances, etc.), these types of observations are comparatively rare. In visible wavelengths, observations at these heights are limited to total eclipses. In the ultraviolet (UV) to extreme UV (EUV) range, very few additional observations have been achieved since the pioneering results of the Ultraviolet Coronagraph Spectrometer (UVCS). Aims. One of the objectives of the Full Sun Imager (FSI) channel of the Extreme Ultraviolet Imager (EUI) on board the Solar Orbiter mission has been to provide very wide field-of-view EUV diagnostics of the morphology and dynamics of the solar atmosphere in temperature regimes that are typical of the lower transition region and of the corona. Methods. FSI carries out observations in two narrowbands of the EUV spectrum centered on 17.4 nm and 30.4 nm that are dominated, respectively, by lines of Fe IX/X (formed in the corona around 1 MK) and by the resonance line of He II (formed around 80 kK in the lower transition region). Unlike previous EUV imagers, FSI includes a moveable occulting disk that can be inserted in the optical path to reduce the amount of instrumental stray light to a minimum. Results. FSI detects signals at 17.4 nm up to the edge of its field of view (7 R ⊙ ), which is about twice further than was previously possible. Operation at 30.4 nm are for the moment compromised by an as-yet unidentified source of stray light. Comparisons with observations by the LASCO and Metis coronagraphs confirm the presence of morphological similarities and differences between the broadband visible light and EUV emissions, as documented on the basis of prior eclipse and space-based observations. Conclusions. The very-wide-field observations of FSI out to about 3 and 7 R ⊙ , without and with the occulting disk, respectively, are paving the way for future dedicated instruments
Ce:LYSO, from scintillator to solid-state lighting as a blue luminescent concentrator
No full textInternational audienc
Low-Latency Masking with Arbitrary Protection Order Based on Click Elements
No full textInternational audienceMasking is the main countermeasure against side-channel attacks due to its sound formal proof of security and the scalability of its protection parameters. However, effective masking increases the implementation complexity by requiring additional silicon area, random number generators and higher latency. Thus, reducing the masking implementation costs while conserving its robustness under side-channel attacks is a relevant branch of research in hardware security applications. Relying on the two-phase bundled-data protocol, this work presents a low-latency masking implementation with arbitrary protection order. In particular, we base our approach on the click elements to control the handshake logic, allowing us to implement asynchronous circuits using conventional synthesis tools. In this manner, we are able to obtain an effective single-cycle and protected implementation of the AES S-box requiring smaller silicon area and potentially lower power consumption compared to the state-of-the-art. Additionally, we detail the asynchronous design methodology that can be applied in different scenarios to improve the latency of secure hardware designs. Finally, we assess leakages to evaluate the robustness of our approach against side-channel attacks
Ultrafast laser-induced plasma anisotropy in pristine and surface pre-structured zinc telluride, probed by terahertz pulses
No full textInternational audienceWe use THz probe pulses to detect and analyze the dynamics of charge transport anisotropies generated by ultrafast laser two-photon absorption in Zinc Telluride (ZnTe) semi-insulating crystal showing smooth and laser structured surfaces. The detected anisotropy consists in a modulation of the THz transmission as a function of the orientation of the axis of ZnTe. The change in THz transmission after pump excitation is attributed to free carrier absorption of the THz field in the laser-induced electron-hole plasma. Pre-structuring the surface sample with laser-induced periodic surface structures (ripples) has strong influence on free carrier THz transmission and its associated anisotropic oscillation. Within the relaxation dynamics of the laser-induced free carriers, two relaxation times have to be considered in order to correctly describe the dynamics, a fast relaxation, of about 50 picoseconds in pristine sample (90 picoseconds in sample pre-structured with ripples), and a slow one, of about 1.5 nanoseconds. A theoretical model based on classical Drude theory and on the dependence of the two-photon absorption coefficient with the crystal orientation and with the laser polarization is used to fit the experimental results
