Conférence

  • Séminaire passé

    26 octobre 2017 par L. Li : Conception de la chaîne d’approvisionnement de l’hydrogène

    Résumé :

    L’hydrogène est considéré depuis longtemps comme un carburant écologique plein d’avenir parce qu’il ne produit aucune émission polluante. Après de grandes avancées technologiques relatives à l’hydrogène ces dernières années, de nouvelles problématiques émergent. Un des grands défis actuels concerne le déploiement des infrastructures associées. Dans ce contexte, le travail de recherche porte sur la conception d’une chaîne d’approvisionnement de l’hydrogène. La première partie du travail a porté sur un état de l’art du domaine et fera l’objet de l’exposé.

  • Séminaire passé

    21 septembre 2017 par H. Ouarnoughi : Autonomic VM placement on hybrid storage system in IaaS cloud

    Résumé :

    IaaS cloud providers offer virtualized resources (CPU, storage, and network) as Virtual Machines (VM). The growth and highly competitive nature of this economy has compelled them to optimize the use of their data centers, in order to offer attractive services at a lower cost. In addition to investments related to infrastructure purchase and cost of use, energy efficiency is a major point of expenditure (2\% of world consumption) and is constantly increasing. Its control represents a vital opportunity. From a technical point of view, the control of energy consumption is mainly based on consolidation approaches. These approaches, which exclusively take into account the CPU use of physical machines (PM) for the VM’s placement, present however many drawbacks. Indeed, recent studies have shown that storage systems and disk I/O represent a significant part of the data center energy consumption (between 14% and 40%).

    In this thesis we propose a new autonomic model for VM placement optimization based on MAPE-K (Monitor, Analyze, Plan, Execute, Knowledge) whereby in addition to CPU, VM I/O and related storage systems are considered. Our first contribution proposes a multilevel VM I/O tracer which overcomes the limitations of existing I/O monitoring tools. In the Analyze step, the collected I/O traces are introduced in a cost model which takes into account the VM I/O profile, the storage system characteristics, and the cloud environment constraints. We also analyze the complementarity between the two main storage classes, resulting in a hybrid storage model exploiting the advantages of each. Indeed, Hard Disk Drives (HDD) represent energy-intensive and inefficient devices compared to compute units. However, their low cost per gigabyte and their long lifetime may constitute positive arguments. Unlike HDD, flash-based Solid-State Disks (SSD) are more efficient and consume less power, but their high cost per gigabyte and their short lifetime (compared to HDD) represent major constraints.
    The Plan phase has initially resulted in an extension of CloudSim to take into account VM I/O, the hybrid nature of the storage system, as well as the implementation of the previously proposed cost model. Secondly, we proposed several heuristics based on our cost model, integrated and evaluated using CloudSim. Finally, we showed that our contribution improves existing approaches of VM placement optimization by a factor of three.

  • Séminaire passé

    01 juin 2017 par J. Guo : A partition-based exact method for graph coloring problem and its application to dynamic resource allocation

    Résumé :

    Graph coloring problem is a well-known traditional NP-complete problem and designing effective exact algorithms for it is still an interesting topic. By analyzing the graph structure, an exact algorithm called TexaCol is proposed which is capable of getting all solutions of the legal k-coloring for a graph G as well as the chromatic polynomial of G. The algorithm includes three steps: the maximal clique decomposition, the suite construction and the vertices coloring. Furthermore, two exact graph coloring algorithms, PexaCol and AexaCol, have been designed, which are able to obtain partial best solutions and all best solutions of G respectively. Finally, on the basis of these static methods, a dynamic graph coloring algorithm is proposed to effectively solve the dynamic cluster resource assignment problem for Device-to-Device networks. The results show that this dynamic algorithm has good performance in resource utilization, runtime and scalability.

  • Séminaire passé

    11 mai 2017 par Z. Peng : Particle Swarm Optimization for Capacitated Location-Routing Problem

    Résumé :

    We propose a particle swarm optimization algorithm to solve the capacitated location-routing problem(CLRP). A new encoding scheme is proposed to represent the solution of the CLRP. During the improvement procedure, several local search methods allow to improve the quality of each solution. Experiment tests are carried on some benchmark instances. Results show that our method is effective and simple.

  • Séminaire passé

    27 avril 2017 par R. Abdallah

    Résumé :

    Les drones d’abord employés et développés par l’armée, ont connu une forte expansion ces dernières années et ont fini par arriver dans le domaine civil. De nombreuses sociétés se sont positionnées sur ce secteur et proposent désormais des services par un seul drone ou plusieurs et des relevés à l’aide de capteurs (photo, vidéo, chimique, sonore…). L’importance d’un vol en formation de flotte de drones se fonde sur une répartition judicieuse des tâches afin qu’ils puissent accomplir plus rapidement leur mission comme dans la stratégie de défense que les drones doivent adopter en cas de danger. Ceci nécessite un niveau élevé de coordination entre les drones rendu possible par un échange continu d’information entre eux et avec leur station de contrôle. Le réseau ad hoc est une solution prometteuse pour faire communiquer les drones entre eux et avec la station sol. Notre objectif est de développer des stratégies de surveillance de zones sensibles par des drones coopératifs équipés de capteurs ainsi que développer l’approche multi drone/ multi cible en partant de l’existant et en proposant des critères de coopération pour l’observation permettant le calcul de trajectoires optimales de la flotte de drones.
    Une des caractéristiques d’une flotte de drone est leur fonctionnement en réseau qu’il faut assurer sa fiabilité (F), sa maintenabilité (M), sa disponibilité (D) et sa sécurité (S). L’association de ces 4 éléments forme le corpus scientifique appelé la sûreté de fonctionnement. En général, l’objectif de la thèse est de développer des méthodes numériques pour évaluer la sûreté de fonctionnement d’une flotte de drones, inspirées des logiques d’organisation et d’autodéfense du le milieu animal.

  • Séminaire récent…

    30 mars 2017 par H. Obeid : Adaptation of Levant’s differentiator based on barrier function.

    Résumé :

    An algorithm for adaptation of Levant’s differentiator gains is proposed for the case when the upper bound of the second derivative exists but it is unknown. The proposed barrier adaptive algorithm can ensure the convergence of the differentiator to some vicinity of the first derivative. Then, a convergence detection criteria is used to estimate this vicinity. The barrier function is presented to adapt both gains of the differentiator thanks to its feature that can maintain the predefined system’s constraint. The properties of the proposed algorithm and other adaptive Levant’s differentiators in the presence of noise are discussed.

  • Séminaire récent…

    12 janvier 2017 par C. Fitouri : Production and predictive maintenance scheduling: the case of job-shop

    Résumé :

    Maintenance is a strategic function in a company. It aims to ensure the proper functioning of an equipment. It no longer has the sole objective to repair a system but also to anticipate and prevent its failures. It aims to reduce the frequency of breakdowns and maximize the operational availability of equipment (minimize the duration of periods of inactivity as a result of voluntary or accidental service interruptions). Our study focuses on the aspects of decision-making. We must know how to determine the appropriate maintenance policy to ensure the operating safety of production tasks allocated on the equipment while optimizing its use. Recently, new failure prediction techniques have emerged. They provide information on the real-time condition of equipment and its remaining useful life. Thus, we can consider new methods of decision-making (allocation of predictive maintenance tasks, production scheduling) based on the analysis of these recent developments.

  • Séminaire récent…

    1er décembre 2016 par A. Chariete (Orange Labs) : Optimisation de la planification des réseaux WLANs en Indoor

    Résumé :

    In this study, we have evaluated several optimization approaches for designing WLAN networks in an Indoor environment. These approaches integrate all relevant parameters to find the optimal placement of access points, there optimum settings (transmission power, azimuth, radiation patterns, etc.) and the optimal allocation of channels while maximizing the coverage, minimizing the interference and minimizing the number of access points to install. Our approaches were evaluated when planning a WLAN 802.11n/ac with a 5GHz band and 20MHz bandwidth on multistory buildings. The results show the effectiveness of our approach to quickly plan a high quality WLANs.