Publications

2022

• Impacts of Feedback Current Value and Learning Rate on Equilibrium Propagation Performance
  
  • Kiraz Fatma Zulal
  • Pham Dang-Kien Germain
  • Desgreys Patricia
  , 2022, pp.519-523. The use of the Equilibrium Propagation algorithm for analog neural networks was introduced in 2020 as an alternative to Backpropagation [1]. In the existing works of analog implementation of Equilibrium Propagation, the impacts of the learning rate, alpha (α), and the scaling factor of the feedback current, beta (β), have not been discussed. This work analyzes the impacts of the scaling factor of feedback current and the learning rate together with the ratio of those two parameters on the algorithm convergence. An Equilibrium Propagation circuit has been implemented on Cadence Virtuoso for a simple task to test the impacts of alpha (α) and beta (β) parameters. Numerical simulations are carried out in a Python-Spectre interface that we implemented. Detecting the optimum ranges for alpha (α) and beta (β) values is particularly influential on the algorithm performance. Our simulation results show that the algorithm only converges for distinctive alpha (α) and beta (β) values. For the tasks we experimented with, the scaling factor should be smaller than 0.1 (β < 0.1), and the learning rate should be larger than 5 × 10−6 and smaller than 0.1 (5 × 10−6 ≤ α < 0.1) to make the algorithm converge. (10.1109/NEWCAS52662.2022.9842178)
  DOI : 10.1109/NEWCAS52662.2022.9842178
• Worst-case Analysis for Interactive Evaluation of Boolean Provenance
  
  • Amarilli Antoine
  • Amsterdamer Yael
  , 2022. In recent work, we have introduced a framework for fine-grained consent management in databases, which combines Boolean data provenance with the field of interactive Boolean evaluation. In turn, interactive Boolean evaluation aims at unveiling the underlying truth value of a Boolean expression by frugally probing the truth values of individual values. The required number of probes depends on the Boolean provenance structure and on the (a-priori unknown) probe answers. Prior work has analyzed and aimed to optimize the expected number of probes, where expectancy is with respect to a probability distribution over probe answers. This paper gives a novel worst-case analysis for the problem, inspired by the decision tree depth of Boolean functions. Specifically, we introduce a notion of evasive provenance expressions, namely expressions, where one may need to probe all variables in the worst case. We show that read-once expressions are evasive, and identify an additional class of expressions (acyclic monotone 2-DNF) for which evasiveness may be decided in PTIME. As for the more general question of finding the optimal strategy, we show that it is coNP-hard in general. We are still able to identify a sub-class of provenance expressions that is "far from evasive", namely, where an optimal worst-case strategy probes only log(n) out of the n variables in the expression, and show that we can find this optimal strategy in polynomial time.
• Efficient enumeration algorithms for annotated grammars
  
  • Amarilli Antoine
  • Jachiet Louis
  • Muñoz Martín
  • Riveros Cristian
  , 2022. We introduce annotated grammars, an extension of context-free grammars which allows annotations on terminals. Our model extends the standard notion of regular spanners, and has the same expressiveness as the extraction grammars recently introduced by Peterfreund. We study the enumeration problem for annotated grammars: fixing a grammar, and given a string as input, enumerate all annotations of the string that form a word derivable from the grammar. Our first result is an algorithm for unambiguous annotated grammars, which preprocesses the input string in cubic time and enumerates all annotations with output-linear delay. This improves over Peterfreund's result, which needs quintic time preprocessing to achieve this delay bound. We then study how we can reduce the preprocessing time while keeping the same delay bound, by making additional assumptions on the grammar. Specifically, we present a class of grammars which only have one derivation shape for all outputs, for which we can enumerate with quadratic time preprocessing. We also give classes generalizing regular spanners for which linear time preprocessing suffices. (10.1145/3517804.3526232)
  DOI : 10.1145/3517804.3526232
• On Efficient and Secure Code-based Masking: A Pragmatic Evaluation
  
  • Wu Qianmei
  • Cheng Wei
  • Guilley Sylvain
  • Zhang Fan
  • Fu Wei
  IACR Transactions on Cryptographic Hardware and Embedded Systems, IACR, 2022, 2022 (3), pp.192-222. Code-based masking is a highly generalized type of masking schemes, which can be instantiated into specific cases by assigning different encoders. It captivates by its side-channel resistance against higher-order attacks and the potential to withstand fault injection attacks. However, similar to other algebraically-involved masking schemes, code-based masking is also burdened with expensive computational overhead. To mitigate such cost and make it efficient, we contribute to several improvements to the original scheme proposed by Wang et al. in TCHES 2020. Specifically, we devise a computationally friendly encoder and accordingly accelerate masked gadgets to leverage efficient implementations. In addition, we highlight that the amortization technique introduced by Wang et al. does not always lead to efficient implementations as expected, but actually decreases the efficiency in some cases.From the perspective of practical security, we carry out an extensive evaluation of the concrete security of code-based masking in the real world. On one hand, we select three representative variations of code-based masking as targets for an extensive evaluation. On the other hand, we aim at security assessment of both encoding and computations to investigate whether the state-of-the-art computational framework for code-based masking reaches the security of the corresponding encoding. By leveraging both leakage assessment tool and side-channel attacks, we verify the existence of “security order amplification” in practice and validate the reliability of the leakage quantification method proposed by Cheng et al. in TCHES 2021. In addition, we also study the security decrease caused by the “cost amortization” technique and redundancy of code-based masking. We identify a security bottleneck in the gadgets computations which limits the whole masked implementation. To the best of our knowledge, this is the first time that allows us to narrow down the gap between the theoretical security order under the probing model (sometimes with simulation experiments) and the concrete side-channel security level of protected implementations by code-based masking in practice. (10.46586/tches.v2022.i3.192-222)
  DOI : 10.46586/tches.v2022.i3.192-222
• Sources pré-déformées pour des analyses haute résolution de l'impact du rayonnement radiofréquence sur l'activité cérébrale
  
  • Henry Clément
  • Merlini Adrien
  • Wiart Joe
  • Andriulli F.P.
  • Person Christian
  , 2022. Cet article présente une approche originale de mesure de signaux EEG d'un patient en présence d'ondes radiofréquences induites dans le cerveau comparables à celle d'un téléphone portable. Outre la quantification de la déformation inhérente à la présence des 256 électrodes d'un casque EEG, l'idée réside dans la génération d'une distribution des champs électromagnétiques au sein des tissus cérébraux conformes à celle que provoque un téléphone mobile en l'absence des électrodes. Nous proposons ainsi une technique de substitution de la source initiale (téléphone mobile générique) par un ensemble de sources équivalentes reconstruisant fidèlement les conditions d'exposition sans électrodes.
• Réalisation et mesure d’un absorbant large bande à métamatériaux avec des matériaux composites structuraux
  
  • Begaud Xavier
  • Lepage Anne Claire
  • Rance Olivier
  • Soiron Michel
  • Barka André
  • Parneix Patrick
  • Laybros Sarah
  , 2022. Cette contribution présente la réalisation et la mesure d’un absorbant à métamatériaux initialement conçu avec des matériaux RF et remplacés par des matériaux composites structuraux, i.e. renforcés de fibres. L’absorbant présente un module du coefficient de réflexion mesuré en incidence normale inférieur à - 13.2 dB de 5.2 GHz à 18 GHz, pour une épaisseur totale de 8.9 mm. Le module de coefficient de réflexion mesuré est inférieur à -10 dB jusqu’à 45° d’incidence quelle que soit la polarisation entre 4.9 et 18 GHz.
• Conception et optimisation d’une structure de type Squarax pour la combinaison spatiale dans la bande de fréquence 2-18 GHz
  
  • Spillebout Théo
  • Bergeault Eric
  • Geslin Florian
  • Gerfault Bertrand
  • Lepage Anne Claire
  • Begaud Xavier
  • Belluot James
  • Leggieri Alberto
  , 2022. Nous présentons les résultats obtenus en simulation pour une architecture Squarax de combinaison spatiale de puissance de 2 à 18 GHz. L’objectif est de démontrer la faisabilité d’un combineur de puissance innovant pour additionner la puissance fournie par les amplificateurs (PA) afin d’atteindre un niveau en sortie de 100 W tout en gardant un niveau de pertes en division et en combinaison inférieur à - 1 dB.
• A Coq Framework for More Trustworthy DRAM Controllers
  
  • Lisboa Malaquias Felipe
  • Asavoae Mihail
  • Brandner Florian
  , 2022, pp.140-150. In order to prove conformance to memory standards and bound memory access latency, recently proposed real-time DRAM controllers rely on paper and pencil proofs, which can be troubling: they are difficult to read and review, they are often shown only partially and/or rely on abstractions for the sake of conciseness, and they can easily diverge from the controller implementation, as no formal link is established between both. We propose a new framework written in Coq, in which we model a DRAM controller and its expected behaviour as a formal specification. The trustworthiness in our solution comes two-fold: 1) proofs that are typically done on paper and pencil are now done in Coq and thus certified by it’s kernel, and 2) the reviewer’s job develops into making sure that the formal specification matches the standards – instead of performing a thorough check of the underlying mathematical formalism. Our framework provides a generic DRAM model capturing a set of controller properties as proof obligations, which all implementations must comply with. We focus on properties related to the respect of timing constraints imposed by the memory standards, the correctness of the DRAM command protocol and the assertiveness that every incoming request is handled in bounded time. We refine our specification with two implementations based on widely-known arbitration policies – First-in First-Out (FIFO) and Time-Division Multiplexing (TDM). We extract proved code from our model and use it as a “trusted core” on a cycle-accurate DRAM simulator. (10.1145/3534879.3534907)
  DOI : 10.1145/3534879.3534907
• RTNS 2022: The 30th International Conference on Real-Time Networks and Systems
  
  • Abdeddaïm Yasmina
  • Cucu-Grosjean Liliana
  • Nelissen Geoffrey
  • Pautet Laurent
  , 2022. (10.1145/3534879)
  DOI : 10.1145/3534879
• Material Reflection Measurements in Centimeter and Millimeter Wave Ranges for 6G Wireless Communications
  
  • Aliouane Mohamed
  • Conrat Jean-Marc
  • Cousin Jean-Christophe
  • Begaud Xavier
  , 2022, pp.43-48. Research in the field of telecommunication is moving towards 6G. Upper mm Wave frequencies (100–300 GHz) are seen as a promising band for 6G. ITU recommendations on material characteristics are limited to 100 GHz. This study aims to further extend the currently existing knowledge on wave-material interaction and make the link between below and above 100 GHz frequencies by providing reflection measurements from 2 to 170 GHz. The material reflection coefficients of seventeen different materials e.g. glass, wood, aerated concrete, and polystyrene are continuously reported along the measured band. The measurement setup is based on a vector network analyzer with millimeter wave extension modules connected starting from 50 GHz. (10.1109/EuCNC/6GSummit54941.2022.9815763)
  DOI : 10.1109/EuCNC/6GSummit54941.2022.9815763
• Side-channel information leakage of code-based masked implementations
  
  • Cheng Wei
  • Rioul Olivier
  • Liu Yi
  • Béguinot Julien
  • Guilley Sylvain
  , 2022. Side-channel attacks (SCAs) are among the most powerful physical attacks against cryptographic implementations. To thwart SCAs, a well-established countermeasure is random masking. A recent code-based masking formalism unifies several known masking schemes and allows one to carry out an all-in-one leakage quantification. In this paper, we investigate how a code-based masked imple- mentation leaks in an information-theoretic setting, where the mutual information measures the impact of both number and positions of probes in the probing attack model. We also establish that the mutual information decreases as the measurement noise variance increases, with an exponent equal to the dual distance of the masking code. Our findings quantitatively connect the attacker’s capability to recover secret keys with the actual mutual information leakage of the protected implementation. (10.1109/CWIT55308.2022.9817673)
  DOI : 10.1109/CWIT55308.2022.9817673
• Estimation de paramètres de resynthèse de sons d'instruments de musique avec des outils de morphologie mathématique
  
  • Romero-García Gonzalo
  • Agón Carlos
  • Bloch Isabelle
  , 2022. Les sons représentés par des spectrogrammes peuvent être considérés comme des images dont la dimension horizontale correspond au temps, et la verticale à la fréquence. Dans cet article, nous proposons d’explorer quelques outils algébriques de la morphologie mathématique, théorie largement développée en analyse et interprétation d’images, pour estimer des paramètres de resynthèse d’un son d’instrument de musique en le modélisant comme une partie harmonique à laquelle est ajouté un bruit blanc filtré. En particulier, nous montrons que des transformations non linéaires visant à détecter des structures saillantes permettent de déduire les paramètres des composantes harmoniques d’un signal. Avec les mêmes outils, nous estimons les paramètres de filtrage pour resynthétiser la partie non-harmonique. (10.5281/zenodo.6800838)
  DOI : 10.5281/zenodo.6800838
• Unsupervised Network Intrusion Detection System for AVTP in Automotive Ethernet Networks
  
  • Alkhatib Natasha
  • Mushtaq Maria
  • Ghauch Hadi
  • Danger Jean-Luc
  , 2022, pp.1731-1738. (10.1109/IV51971.2022.9827285)
  DOI : 10.1109/IV51971.2022.9827285
• When Can We Answer Queries Using Result-Bounded Data Interfaces?
  
  • Amarilli Antoine
  • Benedikt Michael
  Logical Methods in Computer Science, Logical Methods in Computer Science Association, 2022. We consider answering queries on data available through access methods, that provide lookup access to the tuples matching a given binding. Such interfaces are common on the Web; further, they often have bounds on how many results they can return, e.g., because of pagination or rate limits. We thus study result-bounded methods, which may return only a limited number of tuples. We study how to decide if a query is answerable using result-bounded methods, i.e., how to compute a plan that returns all answers to the query using the methods, assuming that the underlying data satisfies some integrity constraints. We first show how to reduce answerability to a query containment problem with constraints. Second, we show "schema simplification" theorems describing when and how result-bounded services can be used. Finally, we use these theorems to give decidability and complexity results about answerability for common constraint classes.
• On the convergence of stochastic primal-dual hybrid gradient
  
  • Alacaoglu Ahmet
  • Fercoq Olivier
  • Cevher Volkan
  SIAM Journal on Optimization, Society for Industrial and Applied Mathematics, 2022, 32 (2), pp.1288-1318. In this paper, we analyze the recently proposed stochastic primal-dual hybrid gradient (SPDHG) algorithm and provide new theoretical results. In particular, we prove almost sure convergence of the iterates to a solution and linear convergence with standard step sizes, independent of strong convexity constants. Our assumption for linear convergence is metric subregularity, which is satisfied for smooth and strongly convex problems in addition to many nonsmooth and/or nonstrongly convex problems, such as linear programs, Lasso, and support vector machines. In the general convex case, we prove optimal sublinear rates for the ergodic sequence and for randomly selected iterate, without bounded domain assumptions. We also provide numerical evidence showing that SPDHG with standard step sizes shows favorable and robust practical performance against its specialized strongly convex variant SPDHG-$\mu$ and other state-of-the-art algorithms including variance reduction methods and stochastic dual coordinate ascent. (10.1137/19M1296252)
  DOI : 10.1137/19M1296252
• Is on-line handwriting gender-sensitive? what tells us a combination of statistical and machine learning approaches
  
  • Likforman-Sulem Laurence
  • Cordasco Gennaro
  • Esposito Anna
  , 2022, 13363. Handwriting is an everyday life human activity. It can be collected off-line by scanning sheets of paper. The resulting images can then be processed by a computer-based system. Thanks to digitizing tablets, handwriting can also be collected on-line. From the collected raw signals (pen position, pressure over time), the dynamics of the writing can be recovered. Since handwriting is unique for each individual, it can be considered as a biometric modality. Biometric systems predicting gender from off-line handwriting, have been recently proposed. However we observe that, in contrast to other modalities such as speech, it is not straightforward for a human being (even expert) to predict gender. In this study we explore the limits of automatic gender prediction from on-line handwriting collected from a young adults population, homogeneous in terms of age and education. In our previous work [1], a statistical analysis of on-line dynamic features has shown differences between male and female groups. In the present study, we provide these features to a classifier, based on a machine learning approach (SVMs). Since datasets are relatively small (240 subjects), several evaluation frameworks are explored: cross validation (CV), bootstrap, and fixed train/test partitions. Accuracies obtained from fixed partitions range from 37% to 79%, while those estimated by CV and bootstrap are around 60%. This shows to our opinion the limits of the gender recognition task from on-line handwriting, for our observed young adult population. (10.1007/978-3-031-09037-0_24)
  DOI : 10.1007/978-3-031-09037-0_24
• Modeling of Amplitude Squeezing in a Pump-Noise-Suppressed Interband Cascade Laser
  
  • Zhao Shiyuan
  • Grillot Frédéric
  IEEE Photonics Journal, Institute of Electrical and Electronics Engineers (IEEE), 2022, 14 (3), pp.1 - 8. A semi-classical rate equation model for the amplitude squeezing of mid-infrared interband cascade lasers is presented based on Green's function method. We analytically investigate the squeezing performance through a small signal analysis, where an additional partition noise in the output power is considered because of the anti-correlation between photons reflected and transmitted at the output laser facet. Thus, the squeezing effect can be achieved in the low frequency region with noise-suppressed pump. Amplitude squeezed states in interband cascade lasers can qualitatively be evaluated depending on the pump noise conditions. Such ultra-low amplitude noise laser sources enable various applications such as high precision measurements, metrology, and spectroscopy range where squeezed states can be used to replace current shot-noise limited laser mid-infrared sources. (10.1109/jphot.2022.3163546)
  DOI : 10.1109/jphot.2022.3163546
• Image Restoration for Remote Sensing: Overview and toolbox
  
  • Rasti Behnood
  • Chang Yi
  • Dalsasso Emanuele
  • Denis Loic
  • Ghamisi Pedram
  IEEE geoscience and remote sensing magazine, IEEE, 2022, 10 (2), pp.201-230. (10.1109/MGRS.2021.3121761)
  DOI : 10.1109/MGRS.2021.3121761
• Complexity of the unicity of an optimal vertex cover in graphs
  
  • Hudry Olivier
  • Lobstein Antoine
  , 2022.
• Teaching and Learning Programming with Linux using In-Browser Client-Side Web Technologies: Exploring the Key Features for Achieving Systems and Tools Scalability
  
  • Sharrock Rémi
  • Collin Ewen
  • Labat Thibaud
  • Hamonic Ella
  • Bonfert-Taylor Petra
  • Goudzwaard Michael
  , 2022, pp.427-430. This paper explores the key features for achieving systems and tools scalability for online learning. We created large scale tools used in popular online courses on teaching and learning programming with Linux that take advantage of client-side web technologies. We advocate for more compatible, interoperable and light tools and systems that are resilient, elastic, traceable, responsive and secure. We also think social features should be considered like friendliness, accessibility, openness and re-usability. Those features could be used to assess the quality and the scalability potential of MOOCs by choosing different metrics to evaluate for each introduced feature. In doing so we hope to diversify the standard indicators of the success of online courses, especially regarding their scalability, and give some examples for STEM courses. (10.1145/3491140.3528295)
  DOI : 10.1145/3491140.3528295
• Survey on smart homes: Vulnerabilities, risks, and countermeasures
  
  • Hammi Badis
  • Zeadally Sherali
  • Khatoun Rida
  • Nebhen Jamel
  Computers & Security, Elsevier, 2022, 117, pp.102677. (10.1016/j.cose.2022.102677)
  DOI : 10.1016/j.cose.2022.102677
• Automatic Support for Requirements Validation
  
  • Yasmine Assioua
  • Ameur-Boulifa Rabéa
  • Guitton-Ouhamou Patricia
  • Pacalet Renaud
  , 2022. The automotive industry is currently going through rapid changes from a mechanical industry to one driven by innovation in electronics and embedded software. This significant change creates also significant challenges to the industry. One of the most important is the ability to create safe vehicles, emphasizing the importance of safety by design. This paper is intended to contribute to current activities working towards an industry-wide development of reliable and secure systems. Correct by design methodology, including formal methods, have the potential to improve dependability of systems in this domain. And their use at an early stage of the development process ensures faster time to market. In this paper, we present tool support for our approach that aims at integrating the formal analysis and verification of functional requirements from early stages of the development life cycle, by using model checking technique. From informal requirement specifications the tool delivers models. They will be used to produce evidences that the requirement specifications are realizable, otherwise it can guide their revision. The approach is illustrated by a case study based on a specific function of autonomous vehicles.
• PCA-AE: Principal Component Analysis Autoencoder for Organising the Latent Space of Generative Networks
  
  • Pham Chi-Hieu
  • Ladjal Saïd
  • Newson Alasdair
  Journal of Mathematical Imaging and Vision, Springer Verlag, 2022, 64 (5), pp.569-585. Autoencoders and generative models produce some of the most spectacular deep learning results to date. However, understanding and controlling the latent space of these models presents a considerable challenge. Drawing inspiration from principal component analysis and autoencoders, we propose the Principal Component Analysis Autoencoder (PCA-AE). This is a novel autoencoder whose latent space verifies two properties. Firstly, the dimensions are organised in decreasing importance with respect to the data at hand. Secondly, the components of the latent space are statistically independent. We achieve this by progressively increasing the latent space during training, and with a covariance loss applied to the latent codes. The resulting autoencoder produces a latent space which separates the intrinsic attributes of the data into different components of the latent space, in a completely unsupervised manner. We also describe an extension of our approach to the case of powerful, pre-trained GANs. We show results on both synthetic examples of shapes and on a state-of-the-art GAN. For example, we are able to separate the colour shade scale of hair, pose of faces and gender, without accessing any labels. We compare the PCA-AE with (10.1007/s10851-022-01077-z)
  DOI : 10.1007/s10851-022-01077-z
• An OPC UA PubSub Implementation Approach for Memory-Constrained Sensor Devices
  
  • Nguyen Quang-Duy
  • Bellot Patrick
  • Petton Pierre-Yves
  IEEE Transactions on Industrial Informatics, Institute of Electrical and Electronics Engineers, 2021. Open Platform Communications Unified Architecture (OPC UA) comprises 14 specifications to deploy an industrial system with reliability, security, and interoperability. While realizing this standard, the devices of the considering industrial system must have enough capabilities in computation and storage. It is a challenge in the Industrial Internet of Things (IIoT), a scenario in which field-level devices can be memory-constrained sensor devices. Tailoring OPC UA to fit such devices requires advanced programming skills; otherwise, system developers must simplify or remove some essential features of OPC UA. This paper presents another implementation approach to tailor OPC UA PubSub, a specification for the publish-subscribe messaging pattern, into memory-constrained sensor devices. This approach, titled OPC UA PubSub-C, proposes using a remote OPC UA server as a configurator to operate large-memory-footprint tasks for field-level devices.
• A Combined Stochastic and Physical Framework for Modeling Indoor 5G Millimeter Wave Propagation
  
  • Nassif Georges
  • Gloaguen Catherine
  • Martins Philippe
  IEEE Transactions on Antennas and Propagation, Institute of Electrical and Electronics Engineers, 2022, 70 (6), pp.4712-4727. (10.1109/TAP.2022.3161286)
  DOI : 10.1109/TAP.2022.3161286