Publications

2019

• Analyse d'opinion dans les interactions orales
  
  • Barrière Valentin
  , 2019. La reconnaissance des opinions d'un locuteur dans une interaction orale est une étape cruciale pour améliorer la communication entre un humain et un agent virtuel. Dans cette thèse, nous nous situons dans une problématique de traitement automatique de la parole (TAP) sur les phénomènes d'opinions dans des interactions orales spontanées naturelles. L'analyse d'opinion est une tâche peu souvent abordée en TAP qui se concentrait jusqu'à peu sur les émotions à l'aide du contenu vocal et non verbal. De plus, la plupart des systèmes récents existants n'utilisent pas le contexte interactionnel afin d'analyser les opinions du locuteur. Dans cette thèse, nous nous penchons sur ces sujet. Nous nous situons dans le cadre de la détection automatique en utilisant des modèles d’apprentissage statistiques. Après une étude sur la modélisation de la dynamique de l'opinion par un modèle à états latents à l’intérieur d'un monologue, nous étudions la manière d’intégrer le contexte interactionnel dialogique, et enfin d'intégrer l'audio au texte avec différents types de fusion. Nous avons travaillé sur une base de données de Vlogs au niveau d'un sentiment global, puis sur une base de données d'interactions dyadiques multimodales composée de conversations ouvertes, au niveau du tour de parole et de la paire de tours de parole. Pour finir, nous avons fait annoté une base de données en opinion car les base de données existantes n'étaient pas satisfaisantes vis-à-vis de la tâche abordée, et ne permettaient pas une comparaison claire avec d'autres systèmes à l'état de l'art.A l'aube du changement important porté par l’avènement des méthodes neuronales, nous étudions différents types de représentations: les anciennes représentations construites à la main, rigides mais précises, et les nouvelles représentations apprises de manière statistique, générales et sémantiques. Nous étudions différentes segmentations permettant de prendre en compte le caractère asynchrone de la multi-modalité. Dernièrement, nous utilisons un modèle d'apprentissage à états latents qui peut s'adapter à une base de données de taille restreinte, pour la tâche atypique qu'est l'analyse d'opinion, et nous montrons qu'il permet à la fois une adaptation des descripteurs du domaine écrit au domaine oral, et servir de couche d'attention via son pouvoir de clusterisation. La fusion multimodale complexe n'étant pas bien gérée par le classifieur utilisé, et l'audio étant moins impactant sur l'opinion que le texte, nous étudions différentes méthodes de sélection de paramètres pour résoudre ces problèmes.
• White matter multi-resolution segmentation using fuzzy set theory
  
  • Delmonte Alessandro
  • Mercier Corentin
  • Pallud Johan
  • Bloch Isabelle
  • Gori Pietro
  , 2019. The neural architecture of the white matter of the brain, obtained using tractography algorithms, can be divided into different tracts. Their function is, in many cases, still an object of study and might be affected in some syndromes or conditions. Obtaining a reproducible and correct segmentation is therefore crucial both in clinics and in research. However, it is difficult to obtain due to the huge number of fibers and high inter-subject variability. In this paper, we propose to segment and recognize tracts by directly modeling their anatomical definitions, which are usually based on relationships between structures. Since these definitions are mainly qualitative, we propose to model their intrinsic vagueness using fuzzy spatial relations and combine them into a single quantitative score mapped to each fiber. To cope with the high redundancy of tractograms and ease interpretation , we also take advantage of a simplification scheme based on a multi-resolution representation. This allows for an interactive and real-time navigation through different levels of detail. We illustrate our method using the Human Connectome Project dataset and compare it to other well-known white matter segmentation techniques.
• A Content-aware Data-plane for Efficient and Scalable Video Delivery
  
  • Desmouceaux Yoann
  • Enguehard Marcel
  • Nguyen Victor
  • Pfister Pierre
  • Shao Wenqin
  • Vyncke Éric
  , 2019 (10-18). Internet users consume increasing quantities of video content with higher Quality of Experience (QoE) expectations. Network scalability thus becomes a critical problem for video delivery as traditional Content Delivery Networks (CDN) struggle to cope with the demand. In particular, contentawareness has been touted as a tool for scaling CDNs through clever request and content placement. Building on that insight, we propose a network paradigm that provides application-awareness in the network layer, enabling the offload of CDN decisions to the data-plane. Namely, it uses chunk-level identifiers encoded into IPv6 addresses. These identifiers are used to perform networklayer cache admission by estimating the popularity of requests with a Least-Recently-Used (LRU) filter. Popular requests are then served from the edge cache, while unpopular requests are directly redirected to the origin server, circumventing the HTTP proxy. The parameters of the filter are optimized through analytical modeling and validated via both simulation and experimentation with a testbed featuring real cache servers. It yields improvements in QoE while decreasing the hardware requirements on the edge cache. Specifically, for a typical content distribution, our evaluation shows a 22% increase of the hit rate, a 36% decrease of the chunk download-time, and a 37% decrease of the cache server CPU load.
• Recent advances in HARQ communications
  
  • Duhamel Pierre
  • Szczecinski Leszek
  • Ciblat Philippe
  • Bassi Francesca
  , 2019.
• Hybrid III-V on silicon lasers for optical communications
  
  • Gallet Antonin
  , 2019. Photonic integration reduces the size and energy consumption of fiber optic communication systems compared to systems assembled from discrete components. This technology has recently attracted a great interest with the progress of integration on InP and the development of silicon photonics. The latter challenges the integration platform on InP as high-performance and low-cost components can be manufactured in foundries originally developed for microelectronics. Lasers are one of the main parts of transceivers for optical communications. With their integration on the silicon platform, transceivers that include the critical functions of light emission, modulation and detection on the same chip can be made. In the heterogeneous integration platform, components are manufactured in high volumes: several tens or even hundreds of components are produced per wafer. In this thesis, I studied theoretically and experimentally the properties of tunable lasers based on silicon ring resonators, directly modulated distributed feedback lasers and low noise high-quality factor lasers
• Développement Industriel de la Cryptographie Quantique : Défis et Opportunités
  
  • Alleaume Romain
  , 2019.
• Distributed Function Chaining with Anycast Routing
  
  • Wion Adrien
  • Bouet Mathieu
  • Iannone Luigi
  • Conan Vania
  , 2019. Current networks more and more rely on virtualized middle-boxes to flexibly provide security, protocol optimization, and policy compliance functionalities. As such, delivering these services requires that the traffic be steered through the desired sequence of virtual appliances. Current solutions introduce a new logically centralized entity, often called orchestrator, needing to build its own holistic view of the whole network so to decide where to direct the traffic. We advocate that such a centralized orchestration is not necessary and that, on the contrary, the same objectives can be achieved by augmenting the network layer routing so to include the notion of service and its chaining. In this paper, we support our claim by designing such a system called NFV Router. We also present an implementation and an early evaluation, showing that we can easily steer traffic through available resources. The proposed approach offers as well valuable features such as incremental deploya-bility, multi-domain service chaining, failure resiliency, and easy maintenance. (10.1145/3314148.3314355)
  DOI : 10.1145/3314148.3314355
• An Experimental Study of the Treewidth of Real-World Graph Data (Extended Version)
  
  • Maniu Silviu
  • Senellart Pierre
  • Jog Suraj
  , 2019. Treewidth is a parameter that measures how tree-like a relational instance is, and whether it can reasonably be decomposed into a tree. Many computation tasks are known to be tractable on databases of small treewidth, but computing the treewidth of a given instance is intractable. This article is the first large-scale experimental study of treewidth and tree decompositions of real-world database instances (25 datasets from 8 different domains, with sizes ranging from a few thousand to a few million vertices). The goal is to determine which data, if any, can benefit of the wealth of algorithms for databases of small treewidth. For each dataset, we obtain upper and lower bound estimations of their treewidth, and study the properties of their tree decompositions. We show in particular that, even when treewidth is high, using partial tree decompositions can result in data structures that can assist algorithms.
• Voltage-Controlled Magnetic Anisotropy MeRAM Bit-Cell over Event Transient Effects
  
  • Maciel Nilson
  • Crespo Marques Elaine
  • Naviner Lirida
  • Cai Hao
  • Yang Jun
  Journal of Low Power Electronics Applications, 2019, 9 (2), pp.1-14. (10.3390/jlpea9020015)
  DOI : 10.3390/jlpea9020015
• Au calcul bien pesé
  
  • Zayana Karim
  Au fil des maths, APMEP, 2019. Balance électronique, balance trébuchet, balance de Roberval : qu'en faire en classe de mathématique ?
• Additively homomorphic encryption and fragmentation scheme for data aggregation inside unattended wireless sensor networks
  
  • Kapusta Katarzyna
  • Memmi Gérard
  • Noura Hassan
  Annals of Telecommunications - annales des télécommunications, Springer, 2019, 74 (3-4), pp.157-165. (10.1007/s12243-018-0684-x)
  DOI : 10.1007/s12243-018-0684-x
• Traffic Analysis Based on Bluetooth Passive Scanning
  
  • Boudabous Safa
  • Garbiso Julian
  • Leroy Bertrand
  • Clémençon Stéphan
  • Labiod Houda
  , 2019, pp.1-6. (10.1109/VTCSpring.2019.8746452)
  DOI : 10.1109/VTCSpring.2019.8746452
• Comparison of nonlinear frequency division multiplexing and OFDM for optical fiber transmissions
  
  • Gemechu Wasyhun Asefa
  , 2019. Nonlinear effects in optical fiber set the ultimate limit to the channel capacity in long-haul optical transmission systems. Advanced nonlinear compensation techniques such as digital backpropagation (DBP) have been proposed as a solution to overcome the channel capacity crunch. However, given theircomputational complexity, in a practical environment their performance gainremains very limited. This triggered a search for a novel communication system design that takes fiber nonlinearity into consideration. A new nonlinearcommunication method, based on the theory of the inverse spectral transform, has been proposed to overcome the nonlinear capacity crunch. Thismethod, originally proposed by Hasegawa in 1993 and called eigenvalue (ormulti-soliton) communication, is based on the fundamental observation thatthe nonlinear spectrum of an optical signal is invariant (except for a triviallinear phase shift) upon propagation in the fiber channel, as described bythe nonlinear Schrödinger equation (NLSE). This means that if the directspectral transform (also known as nonlinear Fourier transform (NFT)) ofthe received signal can be computed, the eigenvalue spectrum can be fullyrecovered.This thesis focuses on a NFT-based communication technique known as nonlinear frequency division multiplexing (NFDM). The NFDM optical systemis numerically assessed and experimentally demonstrated. First, the structure of the proposed single-polarization NFDM system using the continuousspectrum in the normal dispersion regime is presented. To that end, theNFT of the vector NLSE, or Manakov system, was numerically developed.Based on these algorithms the NFDM method was extended to polarizationdivision multiplexed (PMD) systems, and experimentally validated for thefirst time using the continuous spectrum. Finally, the experiment will bereplicated in the anomalous dispersion regime.Additional numerical studies are presented, in order to investigate the implementation challenges of the proposed NFDM techniques for the continuousspectrum modulation.
• Cybersecurity in networking
  
  • Khatoun Rida
  • Mattos Diogo Menezes Ferrazani
  • Duarte Otto Carlos Muniz Bandeira
  Annals of Telecommunications - annales des télécommunications, Springer, 2019, 74 (3-4), pp.123-124. (10.1007/s12243-019-00706-w)
  DOI : 10.1007/s12243-019-00706-w
• Cyber-security of connected vehicles : contributions to enhance the risk analysis and security of in-vehicle communications
  
  • Karray Khaled
  , 2019. During the last decade, technological advances have made the car more and more connected to the outside world. On the flip side, thistechnological transformation has made modern vehicles subject to advanced cyber attacks. The cyber-physical architectures of automotive systems were not designed with security in mind. With the integration of connected platforms into these cyberphysical systems, the threat landscape has radically changed. Lately, multiple security breaches targeting different car manufacturers have been reported mainly by the scientific community. This makes security a critical concern, with a high impact especially on future autonomous driving. In order to address this gap, rigorous security engineering needs to be integrated into the design process of an automotive system and new protection methods adapted to the specificities of the vehicle systems must be introduced. Threat modeling and risk analysis are essential building blocks of this process. In this context, attack trees proved to be a reasonably good way to model attack steps. Nevertheless, given the diversity of architectures, it can quickly become a burden to draw attack trees for all architectures. This thesis tackles the issues of security of connected vehicles. The proposed approach allows enhancing the threat analysis with the automated generation of attack tree used to assist in the risk assessment step. We also propose novel and efficient protection mechanisms for in-vehicle communication networks capable of coping with existing cyber-physical attacks.
• Active engineering of four-wave mixing spectral correlations in multiband hollow-core fibers
  
  • Cordier Martin
  • Orieux Adeline
  • Debord Benoît
  • Gérôme Frédéric
  • Gorse A.
  • Chafer Matthieu
  • Diamanti Eleni
  • Delaye Philippe
  • Benabid Fetah
  • Zaquine Isabelle
  Optics Express, Optical Society of America - OSA Publishing, 2019, 27 (7), pp.9803-9814. We demonstrate theoretically and experimentally a high level of control of the four-wave mixing process in an inert gas–filled inhibited-coupling guiding hollow-core photonic crystal fiber. The specific multiple-branch dispersion profile in such fibers allows both correlated and separable bi-photon states to be produced. By controlling the choice of gas and its pressure and the fiber length, we experimentally generate various joint spectral intensity profiles in a stimulated regime that is transferable to the spontaneous regime. The generated profiles may cover both spectrally separable and correlated bi-photon states and feature frequency tuning over tens of THz, demonstrating a large dynamic control that will be very useful when implemented in the spontaneous regime as a photon pair source. (10.1364/OE.27.009803)
  DOI : 10.1364/OE.27.009803
• Stochastic reverberation model for uniform and non-diffuse acoustic fields
  
  • Badeau Roland
  , 2019. In a recent research report, we introduced a general stochastic reverberation model that aims to represent the statistical properties of reverberation in a broad variety of acoustic environments. A simplified version of this model, dedicated to the particular case of diffuse (i.e. uniform and isotropic) acoustic fields, omnidirectional sources and microphones, and constant attenuation w.r.t frequency, has been investigated both mathematically and experimentally in a recent research paper. We showed that this model provides a common mathematical framework that unifies several well-known results regarding the statistical properties of reverberation in the space, time and frequency domains. In this research report, we aim to extend this mathematical analysis to uniform and non-diffuse acoustic fields, and directive sources and microphones. We show that the predictions of the general stochastic model experimentally match the observations, based on both synthetic and real room impulse responses, measured in various acoustic environments.
• Autoencoding any Data through Kernel Autoencoders
  
  • Laforgue Pierre
  • d'Alché-Buc Florence
  • Clémençon Stéphan
  , 2019.
• A convex optimization framework for video quality and resolution enhancement from multiple descriptions
  
  • Purica Andrei
  • Boyadjis Benoit
  • Pesquet-Popescu Béatrice
  • Dufaux Frédéric
  • Bergeron Cyril
  IEEE Transactions on Image Processing, Institute of Electrical and Electronics Engineers, 2019, 28 (4), pp.1661-1674. Transmission and compression technologies advancement over the past decade led to a shift of multimedia content towards cloud systems. Multiple copies of the same video are available through numerous distribution systems. Different compression levels, algorithms and resolutions are used to match the requirements of particular applications. As 4k display technologies are rapidly adopted, resolution enhancement algorithms are of vital importance. Current solutions do not take into account the particularities of different video encoders, while video reconstruction methods from compressed sources do not provide resolution enhancement. In this paper, we propose a multi source compressed video enhancement framework where each description can have a different compression level and resolution. Using a variational formulation based on a modern proximal dual splitting algorithm, we efficiently combine multiple descriptions of the same video. Two applications are proposed: combining two compressed Low Resolution (LR) descriptions of a video sequence into a High Resolution (HR) description and enhancing a compressed HR video using a LR compressed description. Tests are performed over multiple video sequences encoded with High Efficiency Video Coding (HEVC), at different compression levels and resolutions obtained through multiple down-sampling methods. (10.1109/TIP.2018.2880567)
  DOI : 10.1109/TIP.2018.2880567
• Optimal Trajectories of a UAV Base Station Using Lagrangian Mechanics
  
  • Coupechoux Marceau
  • Darbon Jérôme
  • Kélif Jean-Marc
  • Sigelle Marc
  , 2019, pp.1-6. This paper considers the problem of optimizing the trajectory of an Unmanned Aerial Vehicle (UAV) Base Station (BS). A map is considered, characterized by a traffic intensity of users to be served. The UAV BS must travel from a given initial location at an initial time to a final position within a given duration and serves the traffic on its way. The problem consists in finding the optimal trajectory that minimizes a certain cost depending on the velocity and on the amount of served traffic. The problem is formulated using the framework of Lagrangian mechanics. When the traffic intensity is quadratic (single-phase), we derive closed-form formulas for the optimal trajectory. When the traffic intensity is bi-phase, necessary conditions of optimality are provided and an Alternating Optimization Algorithm is proposed, that returns a trajectory satisfying these conditions. The Algorithm is initialized with a Model Predictive Control (MPC) online algorithm. Numerical results show how the trajectory is improved with respect to the MPC solution.
• DoF of Sectored Cellular Networks with C-RAN
  
  • Gelincik Samet
  • Rekaya-Ben Othman Ghaya
  , 2019.
• A Statistical Model for On-Body Channels in Indoor Considering Rooms Geometry and Subject Location
  
  • Youssef Badre
  • Berrah Manel
  • Roblin Christophe
  , 2019.
• Towards Safety and Security Co-engineering: Challenging Aspects for a Consistent Intertwining
  
  • Pedroza Gabriel
  , 2019, pp.3-16. The emergence of systems identified as both safety and security critical has motivated research and industry communities to search for novel approaches to conduct multi-concern engineering (co-engineering). But several aspects and issues have arisen during the process what has limited the advances. Among them, there are the specificities found in concepts, methods and development cycles, the current standalone practices of safety and security, and the lack of consolidated metrics for safety-security assessment. This paper presents synthetic discussions on referred topics along with some suggestions for solutions and perspectives.
• Propagation channel modeling at centimeter–and–millimeter–wave frequencies in 5G urban micro–cellular context
  
  • Diakhate Cheikh
  , 2019. The advent of bandwidth–demanding mobile applications and services has led to a massive explosion of the network data traffic. In order to alleviate this issue, millimeter–Wave communications systems are a promising technology for future 5G systems thanks to the large amount of bandwidth available in this frequency range. However, in order to take full advantage of this technology, knowledge of the radio propagation channel characteristics in these frequency bands is paramount. Therefore, in this thesis, the objective is to study the frequency–dependence of the propagation channel large scale parameters (LSPs), which describe the main channel characteristics. These LSPs include the building penetration losses, the channel delay spread, the channel azimuth spread and the propagation path–loss. The studies are performed thanks to measurement campaigns conducted in Belfort, in typical 5G deployment scenarios such as outdoor–to–indoor and urban outdoor environments, between 3 and 60 GHz.
• Optimisation des réseaux cellulaires pour les communications professionnelles critiques
  
  • Daher Alaa
  , 2019. Les communications professionnelles et critiques sont établies soit entre utilisateurs du secteur de la sûreté publique soit entre acteurs opérants des infrastructures critiques. Du fait des fortes exigences en termes de couverture, de priorité d’accès, de fiabilité et de résilience, sans oublier les services supplémentaires pour les utilisateurs professionnels, ces communications utilisent généralement les technologies PMR (Professional Mobile Radio). Vu la croissance des demandes de services, des changements importants sont attendus dans le domaine de la PMR. Les technologies PMR historiques échouent en effet à fournir des services à débits de données élevés, tels que les services vidéos et le transfert de photos. Ainsi, l’adaptation des technologies utilisées par les opérateurs commerciaux à la PMR apparaît comme une solution prometteuse. D’autre part, la prochaine génération de réseaux cellulaires prévoit une nouvelle variété d’applications et de services, dont les exigences de performances sont extrêmement hétérogènes. Ils se classent en trois groupes: enhanced Mobile BroadBand (eMBB), massive Machine-Type Communications (mMTC) et Ultra-Reliable Low Latency Communications (URLLC). Récemment, les communications critiques ont été classées dans dans la famille URLLC des cas d’usage car elles sont prioritaires par rapport aux autres types de communications dans le réseau.Dans ce contexte, nous concentrons à renforcer la couverture des réseaux radio fournissant des communications de groupe, service essentiel fournit par les technologies PMR, afin de satisfaire les besoins. Tout d’abord, on évalue la performance des transmissions unicast et multicast, c’est à dire, les transmissions Multicast/Broadcast Single Frequency Network (MBSFN) et Single-Cell Point-To-Multipoint (SC-PTM), en termes de qualité radio, d’efficacité spectrale du système et de couverture de cellules, tout en considérant des configurations MBSFN statiques. Puis, nous étudions un modèle analytique pour estimer le Signal to Interference plus Noise Ratio (SINR) dans un réseau MBSFN.En outre, nous proposons un algorithme simple de répétitions sans requête, comme alternative à l’algoritgme Hybrid Automatic Repeat re-Quest (HARQ), afin d’améliorer la couverture du réseau en présence de communications de groupe. En considérant les caractéristiques du canal radio, ainsi que les contraintes de délai de service, nous justifions que notre modèle fournit un important gain par rapport aux algorithmes de répétitions traditionnels.Enfin, on évalue le compromis entre la couverture et la capacité d’un réseau utilisant les transmissions multicast, qui évolue en fonction de la taille du cluster de stations serveuses. On formule alors un problème d’optimisation dont l’objectif est de maintenir une probabilité de blocage acceptable du système, tout en maximisant le SINR moyen du groupe d’utilisateurs. Pour chaque groupe, on choisit le cluster de cellules d’une manière dynamique, en se fondant sur la minimisation d’une fonction sous-modulaire, qui prend en compte le trafic de chaque cellule du réseau à travers certains poids, ainsi que le SINR moyen du groupe. Ces poids sont optimisés au moyen de la méthode Nelder-Mead, dans le but de diriger la probabilité de blocage vers un certain seuil. Les résultats obtenus montrent l’importance du regroupement dynamique des cellules dans l’amélioration de la capacité et la couverture du système.