Publications

2018

• Homodyne detector blinding attack in continuous-variable quantum key distribution
  
  • Qin Hao
  • Kumar Rupesh
  • Makarov Vadim
  • Alleaume Romain
  Physical Review A : Atomic, molecular, and optical physics [1990-2015], American Physical Society, 2018, 012312. We propose an efficient strategy to attack a continuous-variable (CV) quantum key distribution (QKD) system, which we call homodyne detector blinding. This attack strategy takes advantage of a generic vulnerability of homodyne receivers: A bright light pulse sent on the signal port can lead to a saturation of the detector electronics. While detector saturation has already been proposed to attack CV QKD, the attack we study in this paper has the additional advantage of not requiring an eavesdropper to be phase locked with the homodyne receiver. We show that under certain conditions, an attacker can use a simple laser, incoherent with the homodyne receiver, to generate bright pulses and bias the excess noise to arbitrary small values, fully comprising CV QKD security. These results highlight the feasibility and the impact of the detector-blinding attack. We finally discuss how to design countermeasures in order to protect against this attack.
• Multistage single clad 2 μm TDFA with a shared L-band pump source
  
  • Tench Robert
  • Romano Clément
  • Delavaux Jean-Marc
  Applied optics, Optical Society of America, 2018, 57 (21), pp.5948-5955.
• Caractérisation temporelle d'un canal atmosphérique urbain pour les communications optiques sans fil
  
  • Sauvage Chloé
  • Robert Clélia
  • Sorrente Béatrice
  • Erasme Didier
  , 2018. Les télécommunications sans fil sont fragiles sous certaines conditions météo en particulier le brouillard. Nous avons donc cherché à savoir quelle longueur d'onde se transmet le mieux à faible visibilité. La caractérisation d'un canal de propagation nébuleux urbain réel a permis d'extraire un temps caractéristique d'évolution de l'intensité reçue.
• Impact du canal sur les transmissions satellites - sol par optique adaptative en détection cohérente
  
  • Paillier Laurie
  • Conan Jean-Marc
  • Vedrenne Nicolas
  • Le Bidan Raphaël
  • Artaud Géraldine
  • Jaouën Yves
  , 2018, pp.Session O7-A.
• Communications optiques basées sur la transformée de Fourier non-linéaire
  
  • Yousefi Mansoor
  • Jaouën Yves
  • Gemechu Wasyhun A.
  • Song Mengdi
  , 2018, pp.Session O2-B.
• CONTRÔLE DES FLUCTUATIONS BASSES FRÉQUENCES DANS UN LASERÀ CASCADES QUANTIQUES SOUMISÀ DES TEMPÉRATURES CRYOGÉNIQUES
  
  • Spitz O
  • Wu J.
  • Carras M.
  • Wong C W
  • Grillot F.
  , 2018. L'étude du rapport temps de vie des porteurs sur temps de vie des photons est déterminante pour comprendre la dynamique non-linéaire des lasers à cascades quantiques. Les expériences menées dans le cadre de cette étude montre l'influence de la température sur ce rapport et les conséquences sur le domaine chaotique du laser moyen-infrarouge étudié.
• Opening the parallelogram: Considerations on non-Euclidean analogies
  
  • Murena Pierre-Alexandre
  • Cornuéjols Antoine
  • Dessalles Jean-Louis
  , 2018. Analogical reasoning is a cognitively fundamental way of reasoning by comparing two pairs of elements. Several computational approaches are proposed to efficiently solve analogies: among them, a large number of practical methods rely on either a parallelogram representation of the analogy or, equivalently, a model of proportional analogy. In this paper, we propose to broaden this view by extending the parallelogram representation to differential manifolds, hence spaces where the notion of vectors does not exist. We show that, in this context, some classical properties of analogies do not hold any longer. We illustrate our considerations with two examples: analogies on a sphere and analogies on probability distribution manifold.
• Qualifying Causes as Pertinent
  
  • Sileno Giovanni
  • Dessalles Jean-Louis
  , 2018, pp.2488-2493. Several computational methods have been proposed to evaluate the relevance of an instantiated cause to an observed consequence. The paper reports on an experiment to investigate the adequacy of some of these methods as descriptors of human judgments about causal relevance.
• Discriminative Distance-Based Network Indices with Application to Link Prediction
  
  • Chehreghani Mostafa Haghir
  • Bifet Albert
  • Abdessalem Talel
  The Computer Journal, Oxford University Press (UK), 2018, 61 (7), pp.998-1014. In distance-based network indices, the distance between two vertices is measured by the length of shortest paths between them. A shortcoming of this measure is that when it is used in real-world networks, a huge number of vertices may have exactly the same closeness/eccentricity scores. This restricts the applicability of these indices as they cannot distinguish vertices. Furthermore, in many applications, the distance between two vertices not only depends on the length of shortest paths but also on the number of shortest paths between them. In this paper, first we develop a new distance measure, proportional to the length of shortest paths and inversely proportional to the number of shortest paths, that yields discriminative distance-based centrality indices. We present exact and randomized algorithms for computation of the proposed discriminative indices. Then, by performing extensive experiments, we first show that compared with the traditional indices, discriminative indices have usually much more discriminability. Then, we show that our randomized algorithms can very precisely estimate average discriminative path length and average discriminative eccentricity, using only few samples. Then, we show that real-world networks have usually a tiny average discriminative path length, bounded by a constant (e.g. 2). We refer to this property as the tiny-world property. Finally, we present a novel link prediction method that uses discriminative distance to decide which vertices are more likely to form a link in future, and show its superior performance. (10.1093/comjnl/bxy040)
  DOI : 10.1093/comjnl/bxy040
• Automatic Nonverbal Behavior Generation from Image Schemas
  
  • Ravenet Brian
  • Clavel Chloé
  • Pelachaud Catherine
  , 2018. <p>One of the main challenges when developing Embodied Conversational Agents is to give them the ability to autonomously produce meaningful and coordinated verbal and nonverbal behaviors. The relation between these means of communication is more complex than a direct mapping that has often been applied in previous models. In this paper, we propose an intermediate mapping approach we apply on metaphoric gestures first but that could be extended to other representational gestures. Leveraging from previous work in text analysis, embodied cognition and co-verbal behavior production, we introduce a framework articulating speech and metaphoric gesture invariants around a common mental representation: Image Schemas. We establish the components of our framework, detailing the different steps leading to the production of the metaphoric gestures, and we present some preliminary results and demonstrations. We end the paper by laying down the perspectives to integrate, evaluate and improve our model.</p>
• Multi-core Fiber Channel Model and Core Dependent Loss Estimation
  
  • Abousief Akram
  • Rekaya-Ben Othman Ghaya
  • Jaouën Yves
  , 2018, paper SpW1G.3.
• MULOG: A GENERIC VARIANCE-STABILIZATION APPROACH FOR SPECKLE REDUCTION IN SAR INTERFEROMETRY AND SAR POLARIMETRY
  
  • Deledalle Charles-Alban
  • Denis L.
  • Tupin Florence
  , 2018. Speckle reduction is a long-standing topic in SAR data pro- cessing. Continuous progress made in the field of image denoising fuels the development of methods dedicated to speckle in SAR images. Adaptation of a denoising technique to the specific statistical nature of speckle presents variable levels of difficulty. It is well known that the logarithm trans- form maps the intrinsically multiplicative speckle into an additive and stationary component, thereby paving the way to the application of general-purpose image denoising meth- ods to SAR intensity images. Multi-channel SAR images such as obtained in interferometric (InSAR) or polarimetric (PolSAR) configurations are much more challenging. This paper describes MuLoG, a generic approach for mapping a multi-channel SAR image into real-valued images with an additive speckle component that has a variance approxi- mately constant. With this approach, general-purpose image denoising algorithms can be readily applied to restore InSAR or PolSAR data. In particular, we show how recent denois- ing methods based on deep convolutional neural networks lead to state-of-the art results when embedded with MuLoG framework.
• A survey of exemplar-based texture synthesis methods
  
  • Akl Adib
  • Yaacoub Charles
  • Donias Marc
  • da Costa Jean-Pierre
  • Germain Christian
  Computer Vision and Image Understanding, Elsevier, 2018, 172, pp.12-24. (10.1016/j.cviu.2018.04.001)
  DOI : 10.1016/j.cviu.2018.04.001
• On the Effect of Aging in Detecting Hardware Trojan Horses with Template Analysis.
  
  • Karimi Naghmeh
  • Guilley Sylvain
  • Danger Jean-Luc
  , 2018. (10.1109/IOLTS.2018.8474089)
  DOI : 10.1109/IOLTS.2018.8474089
• Hamiltonians for one-way quantum repeaters
  
  • Miatto Filippo M
  • Epping Michael
  • Lutkenhaus Norbert
  Quantum, Verein zur Förderung des Open Access Publizierens in den Quantenwissenschaften, 2018. Quantum information degrades over distance due to the unavoidable imperfections of the transmission channels, with loss as the leading factor. This simple fact hinders quantum communication , as it relies on propagating quantum systems. A solution to this issue is to introduce quantum repeaters at regular intervals along a lossy channel, to revive the quantum signal. In this work we study unitary one-way quantum repeaters, which do not need to perform measurements and do not require quantum memories , and are therefore considerably simpler than other schemes. We introduce and analyze two methods to construct Hamiltonians that generate a repeater interaction that can beat the fundamental repeaterless key rate bound even in the presence of an additional coupling loss, with signals that contain only a handful of photons. The natural evolution of this work will be to approximate a repeater interaction by combining simple optical elements.
• QUALIFE -- Qualité des alliages ferreux : une approche diachronique et statistique
  
  • Gosselin Manon
  • Téreygeol Florian
  • Pagès Gaspard
  • Arribet-Deroin Danielle
  • Tendero Yohann
  • Dillmann Philippe
  , 2018.
• Discovering patterns in high-dimensional extremes
  
  • Chiapino Maël
  , 2018. We present and study unsupervised learning methods of multivariate extreme phenomena in high-dimension. Considering a random vector on which each marginal is heavy-tailed, the study of its behavior in extreme regions is no longer possible via usual methods that involve finite means and variances. Multivariate extreme value theory provides an adapted framework to this study. In particular it gives theoretical basis to dimension reduction through the angular measure. The thesis is divided in two main part: - Reduce the dimension by finding a simplified dependence structure in extreme regions. This step aim at recover subgroups of features that are likely to exceed large thresholds simultaneously. - Model the angular measure with a mixture distribution that follows a predefined dependence structure. These steps allow to develop new clustering methods for extreme points in high dimension.
• Détection dense de changements par réseaux de neurones siamois
  
  • Daudt Rodrigo Caye
  • Le Saux Bertrand
  • Boulch Alexandre
  • Gousseau Yann
  , 2018. Cet article présente d'une part une base de données pu-blique de détection de changements urbains créée à par-tir d'images satellitaires multispectrales Sentinelle-2, et d'autre part des architectures de réseaux neuronaux convo-lutifs pour la détection de changements entre deux images. Les réseaux proposés sont des extensions siamoises d'ar-chitectures entièrement convolutives. Ils sont capables d'apprendre à détecter des changements en utilisant des paires d'images annotées en termes de changement, sans intervention humaine et sans post-traitement. Nous mon-trons leur efficacité tant sur les bases de données RVB de l'état de l'art que sur la nouvelle base multispectrale. En particulier, ces réseaux atteignent de meilleures performances que les méthodes précédemment proposées, tout en étant au moins 500 fois plus rapides que celles-ci. Mots Clef Détection de changements, apprentissage automatique, ré-seaux entièrement convolutifs, observation de la Terre. Abstract This paper presents convolutional neural network architec-tures which perform change detection using a pair of co-registered images. Most notably, we propose Siamese extensions of fully convolutional networks which use heu-ristics about the current problem to achieve the best results in our tests on two open change detection datasets, using both RGB and multispectral images. We show that our system is able to learn from scratch using annotated change detection images. Our architectures achieve better performance than previously proposed methods, while being at least 500 times faster than related systems. We also present a change detection dataset that was developed using Sentinel-2 images.
• Statistical Uplink Dimensioning in Licensed Cellular Low Power IoT Networks
  
  • Mroueh Lina
  • Yu Yi
  • Terré Michel
  • Martins Philippe
  , 2018, pp.527-531. (10.1109/ICT.2018.8464876)
  DOI : 10.1109/ICT.2018.8464876
• Effective capacity based resource allocation for Rayleigh-fading parallel channels
  
  • Ciblat Philippe
  • Stupia Ivan
  • Vandendorpe Luc
  , 2018, pp.1-5. We address the problem of allocating different powers amongst parallel channels when effective capacity is the performance metric and sum-power is constrained. We assume that Chase-Combining-HARQ mechanism is applied. Closed-form expressions for the powers are exhibited. Numerical comparisons with other power allocations obtained through either ergodic capacity or throughput optimizations are done. (10.1109/spawc.2018.8445897)
  DOI : 10.1109/spawc.2018.8445897
• A security vulnerability analysis of SoCFPGA architectures
  
  • Chaudhuri Sumanta
  , 2018, pp.139:1-139:6. SoCFPGAs or FPGAs integrated on the same die with chip multi processors have made it to the market in the past years. In this article we analyse various security loopholes, existing precautions and countermeasures in these architectures. We consider Intel Cyclone/Arria devices and Xilinx Zynq/Ultrascale devices. We present an attacker model and we highlight three different types of attacks namely direct memory attacks, cache timing attacks, and rowhammer attacks that can be used on inadequately protected systems. We present and compare existing security mechanisms in this architectures, and their shortfalls. We present real life example of these attacks and further countermeasures to secure systems based on SoCFPGAs. (10.1145/3195970.3195979)
  DOI : 10.1145/3195970.3195979
• LINKY: un modèle de controverse sur l'innovation contemporaine
  
  • Draetta Laura
  , 2018. Depuis plus de deux ans, les compteurs communicants Linky remplacent peu à peu les anciens compteurs électriques. Une action de modernisation du réseau à l’échelle nationale qui ne se fait pas sans soulever des débats. La controverse autour de ce boîtier compte de nombreuses facettes : des questions de santé publique à celle des données personnelles, de la critique des risques potentiels à celle des modalités de mise en œuvre du projet de déploiement. Pour Laura Draetta, sociologue à Télécom ParisTech, il s’agit là d’un véritable champ d’étude. Linky pose en effet la question de ce qu’est une innovation responsable aujourd’hui, notamment en interrogeant l’importance de la prise en compte des citoyens dans les décisions d’innovation, surtout dans un contexte d’innovation technologique de type infrastructurel aux implications sociales multiples et imprévisibles.
• Divergent-beam backprojection-filtration formula with applications to region-of-interest imaging
  
  • Reshef Aymeric
  • Riddell Cyril
  • Trousset Yves
  • Ladjal Saïd
  • Bloch Isabelle
  , 2018. We propose a new backprojection-filtration (BPF) method for cone-beam computed tomography (CBCT) with flat-panel detectors over circular orbits. The method is exact in the fan-beam geometry and provides an approximate CBCT reconstruction that is different from the standard Feldkamp-Davis-Kress (FDK) method. More interestingly, it can be used for region-of-interest (ROI) reconstruction by complementing a truncated low-noise acquisition with dense angular sampling by additional non-truncated views that are either high-noise or angularly undersampled.
• Recognition Over Encrypted Faces
  
  • Chabanne Hervé
  • Lescuyer Roch
  • Milgram Jonathan
  • Morel Constance
  • Prouff Emmanuel
  , 2019, 11005, pp.174-191. Neural Networks (NN) are today increasingly used in Machine Learning where they have become deeper and deeper to accurately model or classify high-level abstractions of data. Their development however also gives rise to important data privacy risks. This observation motives Microsoft researchers to propose a framework, called Cryptonets. The core idea is to combine simplifications of the NN with Fully Homomorphic Encryptions (FHE) techniques to get both confidentiality of the manipulated data and efficiency of the processing. While efficiency and accuracy are demonstrated when the number of non-linear layers is small (e.g. 2), Cryptonets unfortunately becomes ineffective for deeper NNs which let the privacy preserving problem open in these contexts. This work successfully addresses this problem by combining several new ideas including the use of the batch normalization principle and the splitting of the learning phase in several iterations. We experimentally validate the soundness of our approach with a neural network with 6 non-linear layers. When applied to the MNIST database, it competes with the accuracy of the best non-secure versions, thus significantly improving Cryptonets. Additionally, we applied our approach to secure a neural network used for face recognition. This problem is usually considered much harder than the MNIST hand-written digits recognition and can definitely not be addressed with a simple network like Cryptonets. By combining our new ideas with an iterative (learning) approach we experimentally show that we can build an FHE-friendly network achieving good accuracy for face recognition (10.1007/978-3-030-03101-5_16)
  DOI : 10.1007/978-3-030-03101-5_16
• Questioning the security and efficiency of the ESIoT approach
  
  • Diop Aida
  • Gharout Said
  • Laurent Maryline
  • Leneutre Jean
  • Traoré Jacques
  , 2018, pp.202 - 207. ESIoT was introduced at WiSec 2017 as a protocol for providing secure access control and authentication in Internet of Things (IoT) applications. The core primitive of ESIoT is an identity-based broadcast encryption scheme called Secure Identity-Based Broadcast Encryption (SIBBE). SIBBE is designed to provide secure key distribution among a group of devices in IoT networks, and enable devices in each group to perform mutual authentication. The scheme is also designed to hide the structure of the group from nodes outside of the group. We identify multiple efficiency and security issues in the primitive that prove SIBBE unsuitable for IoT applications. First, we show that the size of the ciphertexts generated by the encryption function is linear in the number of devices in the group as opposed to constant as claimed in the description of the scheme. Additionally, we demonstrate how constrained nodes in the network perform a number of decryptions also linear in the set of devices, implying scalability issues and thus inefficiency for IoT applications. In terms of security, we prove that SIBBE does not achieve the desired property of anonymity and allows an attacker to gain information on the structure of any given group. Finally, we demonstrate how SIBBE does not achieve chosen-ciphertext security as claimed. We however prove its security for a weaker security notion (namely selective-ID indistinguishability against chosen-plaintext attacks) under a strong cryptographic assumption (10.1145/3212480.3212491)
  DOI : 10.1145/3212480.3212491