Dissertations / Theses on the topic 'Equilibrage des charge'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 48 dissertations / theses for your research on the topic 'Equilibrage des charge.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Quintin, Jean-noël. "Equilibrage de charge dynamique sur plates-formes hiérarchiques." Phd thesis, Université de Grenoble, 2011. http://tel.archives-ouvertes.fr/tel-00661447.
Full textQuintin, Jean-Noël. "Equilibrage de charge dynamique sur plates-formes hiérarchiques." Thesis, Grenoble, 2011. http://www.theses.fr/2011GRENM066/document.
Full textThe race towards more processing power between all different hardware manufacturers has in recent years faced deep changes. We see nowadays a huge development in the use of parallel machines with more and more cores and increasingly complex architectures. It seems now clear that we will witness in a near future the development of cheap Network On Chip computers. Executing parallel applications on such machines raises several problems. Amongst them we take in this work interest in the problem of scheduling a set of tasks on a set of computing resources. Between all existing methods we can generally distinguish on-line or off-line algorithms. On-line algorithms like work-stealing present the advantage to work without informations on hardware or tasks durations but do not generally achieve an efficient control of communications. In this book we take interest in on-line tasks scheduling on complex platforms where networking can impact (through congestion) performance. More precisely, we propose several new scheduling algorithms based on work-stealing targeting two different configurations. In a first study, we consider applications whose dependency graph is known in advance. By taking advantage of this information we manage to limit the amount of data transfered and thus to achieve high performance and even outperform the best known off-line algorithms. Concurrently to that, we also study possible optimisations in the case where knowledge of platform topology is available. We show again that it is possible to use this information to enhance performance. Our work allows therefore to extend the application field of scheduling algorithms towards more complex architectures and we hope will allow a better use of tomorrow's machine
Renard, Hélène. "Equilibrage de charge et redistribution de données sur plates-formes hétérogènes." Phd thesis, Ecole normale supérieure de lyon - ENS LYON, 2005. http://tel.archives-ouvertes.fr/tel-00012133.
Full textJuganaru, Mihaela. "Equilibrage et régulation de charge dans les machines parallèles à mémoire distribuée." Phd thesis, Ecole Nationale Supérieure des Mines de Saint-Etienne, 1999. http://tel.archives-ouvertes.fr/tel-00822691.
Full textPhan, Hoang Anh. "Equilibrage de charge et diffusion multicast dans les systèmes pair-à-pair." Paris 7, 2010. http://www.theses.fr/2010PA077123.
Full textWithin a Peer-to-Peer (P2P) Systems a large number of users interact ail together in order to gather ressources in a dynamic environnement. Several decentralized P2P Systems construct an overlay network based on a de Bruijn graph, and offer a good tradeoff between network diameter and node average degree, with a simple routing algorithm. However, they do not efficiently balance the network's load and do not support some more and more popular kinds of communications in the internet such as multicast. This thesis work proposes significant improvements to these Systems, and an efficient multicast algorithm. A first contribution is the improvement of the d-dimensional D2B protocol. First, we describe the protdcol in detail, and then we propose solutions to problems of network coherence (related to its dynamîcity) and to fault tolerance. A second contribution relates to load balancing in such Systems. After the description of the main existing load balancing methods, we present our solutions for balancing the nodes1 degrees. A third and final contribution is dedicated to multicast in a P2P System, at the application level. After a short state of the art of existing methods, we present our concept of "tree-farm". This solution allows for building families of sets of interior-noeud-disjoint trees for efficient multicast communications in P2P Systems based on de Bruijn graphs. The three contributions are evaluated by means of simulations and compared with existing methods of the state of the art
Pierson, Jean-Marc. "Equilibrage de charge dirige par les donnees. Applications a la synthese d'images." Lyon, École normale supérieure (sciences), 1996. http://www.theses.fr/1996ENSL0032.
Full textPérotin, Matthieu. "Calcul haute performance sur matériel générique." Thesis, Tours, 2008. http://www.theses.fr/2008TOUR4022/document.
Full textTwo facts are motivating this work: the demand for High Performance Computing of researchers and the low usage of the computing power of the pedagogic ressources. This thesis aims at giving an answer to the demand for HPC, while preserving the pedagogic ressources for the teaching. This work looked for a solution that would be simple and straightforward for the final users. Their needs and wishes lead to the definition of some specifications, in which most of the constraints could be satisfied with the use of a well designed software stack. Some others, however, cannot be satisfied with the use of existing solutions only, they define a new scheduling problem, in which the goal is to schedule the processes on the available ressources. This problem was studied and solved with various heurisitcs, which performances were compared with a simulator before being implemented in an experimental setup
Renard, Hélène. "Équilibrage de charge et redistribution de données sur plates-formes hétérogènes." Lyon, École normale supérieure (sciences), 2005. http://www.theses.fr/2005ENSL0344.
Full textIn this thesis, we study iterative algorithms onto heterogeneous platforms. These iterative algorithms operate on large data samples (recursive convolution, image processing algorithms, etc. ). At each iteration, independent calculations are carried out in parallel, and some communications take place. An abstract view of the problem is the following: the iterative algorithm repeatedly operates on a large rectangular matrix of data samples. This data matrix is split into vertical (or horizontal) slices that are allocated to the processors. At each step of the algorithm, the slices are updated locally, and then boundary information is exchanged between consecutive slices. This (virtual) geometrical constraint advocates that processors be organized as a virtual ring. Then each processor will only communicate twice, once with its (virtual) predecessor in the ring, and once with its successor. Note that there is no reason a priori to restrict to a uni-dimensional partitioning of the data, and to map it onto a uni-dimensional ring of processors. But uni-dimensional partitionings are very natural for most applications, and, as will be shown in this thesis, the problem to find the optimal one is already very difficult. After dealing with the problems of mapping and load-balancing onto heterogeneous platforms, we consider the problem of redistributing data onto these platforms, an operation induced by possible variations in the resource performances (CPU speed, communication bandwidth) or in the system/application requirements (completed tasks, new tasks, migrated tasks, etc. ). For homogeneous rings the problem has been completely solved. Indeed, we have designed optimal algorithms, and provided formal proofs of correctness, both for unidirectional and bidirectional rings. For heterogeneous rings there remains further research to be conducted. The unidirectional case was easily solved, but the bidirectional case remains open. Still, we have derived an optimal solution for light redistributions, an important case in practice
Haj, Mahmoud Yanal. "Modélisation et évaluation de performances des systèmes de distribution de charge en environnements répartis." Paris 6, 1999. http://www.theses.fr/1999PA066664.
Full textVuchener, Clement. "Equilibrage de charges dynamique avec un nombre variable de processeurs basé sur des méthodes de partitionnement de graphe." Thesis, Bordeaux, 2014. http://www.theses.fr/2014BORD0012/document.
Full textLoad balancing is an important step conditioning the performance of parallel programs. If the workload varies drastically during the simulation, the load must be redistributed regularly among the processors. Dynamic load balancing is a well studied subject but most studies are limited to an initially fixed number of processors. Adjusting the number of processors at runtime allows to preserve the parallel code efficiency or to keep running the simulation when the memory of the current resources is exceeded.In this thesis, we propose some methods based on graph repartitioning in order to rebalance the load while changing the number of processors. We call this problem \M x N repartitioning". These methods are split in two main steps. Firstly, we study the migration phase and we build a \good" migration matrix minimizing several metrics like the migration volume or the number of exchanged messages. Secondly, we use graph partitioning heuristics to compute a new distribution optimizing the migration according to the previous step results. Besides, we propose a direct k-way partitioning algorithm that allows us to improve our biased partitioning. Finally, an experimental study validates our algorithms against state-of-the-art partitioning tools
Hoenen, Olivier. "Parallélisation de méthodes adaptatives semi-Lagrangiennes pour la résolution de l'équation de Vlasov." Université Louis Pasteur (Strasbourg) (1971-2008), 2008. http://www.theses.fr/2008STR13078.
Full textUnderstanding phenomena in plasma physics is an important research area which is associated with several international projects. When numerical simulation of these phenomena is performed by using the kinetic model, charged particles evolution is given by the Vlasov equation. It is a partial differential equation that lies in phase space, which is a 6 dimensional space in the real case. Thus, the numerical resolution of such an equation represents a huge amount of data and computations. In this thesis, we are interested in the efficient implementation of numerical methods based on adaptative meshes. More precisely, we are concerned about their parallelization for distributed memory architectures. We have developped a first parallel solver which uses some highly adaptive mechanisms. We have shown that the overhead of such a solver is associated with the mesh adaptation mechanism and with non predictable communications. Then we have used another mesh adaptation mechanism in order to propose a new block-based solver. For the parallelization of this second solver, priority is given to regularization of both data structure and communications. We have used communication overlapping with computations and dynamic load balancing to make this adaptive solver efficient for a 4 dimensional phase space. Currently, such results were acheived only on shared memory architectures
Sridi, Marwa. "Un modèle de structure de données Cache-aware pour un parallélisme et un équilibrage dynamique de la charge." Thesis, Université Grenoble Alpes (ComUE), 2016. http://www.theses.fr/2016GREAM095.
Full textThe current parallel architectures integrate processors with many cores to shared memory growing and responding to specific usage constraints, particularly in the cache management. To take advantage of this power, a unique distributed memory parallelism, to manage the inter-node communications is not directly adapted to the characteristics of multi-core architectures. In addition, the shared memory computing environments offer techniques for balancing the load among available cores more appropriate than those in a distributed memory context.Thus, programming models like OpenMP and KAAPI is a tailored response to the specific characteristics of these architectures.Given these issues, we are interested in developing a hardware-aware approach taking into consideration the hierarchical organization of parallel architectures with shared memory. Our approach provides an optimization model for the use of storage space in this context of parallelism.To prove the pertinence of this approach, we have implemented it in the fast dynamic simulation software EUROPLEXUS of fluids and structures, focusing on the shared memory parallelism complementary to the distributed memory approach developed elsewhere. Because of its wide range of applications, EUROPLEXUS is characterized by a very rich data structure and very complex dependencies among its routines. We focused on accelerating the main loop iterating over the mesh elements. The heterogeneity of the formulations and the materials of the elements that can co-exist in the same simulated model generates a large variability between the costs of the iterations of this loop. A first parallelization of the loop with the XKAAPI library based on a dynamic workstealing scheduling has been implemented. However, despite the acceleration achieved by the parallel implementation, performance has been restrained by frequent and dispersed access costs to a complex data structure. This makes the implementation of the code difficult to optimize. Because of this structure, much of the execution time has elapsed in cache misses. The work is based on the implementation of a model approximating the data structure that ensures a better access locality. It mainly consists in moving from the global data structure in which the physical fields are stored in separate tables to a structure based on the storage of data in independent structures called groups. These groups contain the data relating to a number of elements in the local tables. This number is an adjustable parameter depending on the size of the cache levels. Specifically, this method returns to the nest of the elementary loop in a loop iterating on groups. The iterations among the groups are distributed over the cores of the architecture.The execution of the inner loop is sequentially by core. The best results are obtained for groups of the L2 cache size. For this particular size, the use of a dynamic load balancing in XKAAPI allowed us to double the acceleration of the elementary loop compared to the reference version of the code. The second part of this thesis is based on the parallelization of elementary loop inside the already parallelized loop. We demonstrated that this second level of parallelism is less efficient than the single. However, this nested parallelism might be interesting on Intel Xeon Phi architectures incorporating hyper-threaded cores at their calculation units
Bamha, Mostafa. "Parallélisme et équilibrage de charges dans le traitement de la jointure et de la multi-jointure sur des architectures SN." Orléans, 2000. http://www.theses.fr/2000ORLE2001.
Full textLe, Pape Cécile. "Contrôle de qualité des données répliquées dans un Cluster." Paris 6, 2005. http://www.theses.fr/2005PA066433.
Full textPredari, Maria. "Load balancing for parallel coupled simulations." Thesis, Bordeaux, 2016. http://www.theses.fr/2016BORD0369/document.
Full textLoad balancing is an important step conditioning the performance of parallel applications. The goal is to distribute roughly equal amounts of computational load across a number of processors, while minimising interprocessor communication. A common approach to model the problem is based on graph structures and graph partitioning algorithms. Moreover, new challenges involve the simulation of more complex physical phenomena, where different parts of the computational domain exhibit different physical behavior. Such simulations follow the paradigm of multi-physics or multi-scale modeling approaches. Combining such different models in massively parallel computations is still a challenge to reach high performance. Additionally, traditional load balancing algorithms are often inadequate, and more sophisticated solutions should be explored. In this thesis, we propose new graph partitioning algorithms that balance the load of such simulations, refered to as co-partitioning. We formulate this problem with the use of graph partitioning with initially fixed vertices which we believe represents efficiently the additional constraints of coupled simulations. We have therefore developed a direct algorithm for graph partitioning that manages successfully problems with fixed vertices. The algorithm is implemented inside Scotch partitioner and a series of experiments were carried out on the DIMACS graph collection. Moreover we proposed three copartitioning algorithms that respect the constraints of the respective coupled codes. We finally validated our algorithms by an experimental study comparing our methods with current strategies on artificial cases and on real-life coupled simulations
Li, Pei. "Unified system of code transformation and execution for heterogeneous multi-core architectures." Thesis, Bordeaux, 2015. http://www.theses.fr/2015BORD0441/document.
Full textHeterogeneous architectures have been widely used in the domain of high performance computing. However developing applications on heterogeneous architectures is time consuming and error-prone because going from a single accelerator to multiple ones indeed requires to deal with potentially non-uniform domain decomposition, inter-accelerator data movements, and dynamic load balancing. The aim of this thesis is to propose a solution of parallel programming for novice developers, to ease the complex coding process and guarantee the quality of code. We lighted and analysed the shortcomings of existing solutions and proposed a new programming tool called STEPOCL along with a new domain specific language designed to simplify the development of an application for heterogeneous architectures. We evaluated both the performance and the usefulness of STEPOCL. The result show that: (i) the performance of an application written with STEPOCL scales linearly with the number of accelerators, (ii) the performance of an application written using STEPOCL competes with an handwritten version, (iii) larger workloads run on multiple devices that do not fit in the memory of a single device, (iv) thanks to STEPOCL, the number of lines of code required to write an application for multiple accelerators is roughly divided by ten
Benkrid, Soumia. "Le déploiement, une phase à part entière dans le cycle de vie des entrepôts de données : application aux plateformes parallèles." Thesis, Chasseneuil-du-Poitou, Ecole nationale supérieure de mécanique et d'aérotechnique, 2014. http://www.theses.fr/2014ESMA0027/document.
Full textDesigning a parallel data warehouse consists of choosing the hardware architecture, fragmenting the data warehouse schema, allocating the generated fragments, replicating fragments to ensure high system performance and defining the treatment strategy and load balancing.The major drawback of this design cycle is its ignorance of the interdependence between subproblems related to the design of PDW and the use of heterogeneous metrics to achieve thesame goal. Our first proposal defines an analytical cost model for parallel processing of OLAP queries in a cluster environment. Our second takes into account the interdependence existing between fragmentation and allocation. In this context, we proposed a new approach to designa PDW on a cluster machine. During the fragmentation process, our approach determines whether the fragmentation pattern generated is relevant to the allocation process or not. The results are very encouraging and validation is done on Teradata. For our third proposition, we presented a design method which is an extension of our work. In this phase, an original method of replication, based on fuzzy logic is integrated
Epimakhov, Igor. "Allocation des ressources pour l'optimisation de requêtes dans les systèmes de grille de données." Toulouse 3, 2013. http://thesesups.ups-tlse.fr/2017/.
Full textData grid systems are utilized more and more due to their storage and computing capacities. One of the main problems of these systems is the resource allocation for SQL query optimization. Recently, the scientific community published numerous approaches and methods of resource allocation, striving to take into account different peculiarities of data grid systems: heterogeneity, instability and large scale. Centralized management structure predominates in the proposed methods, in spite of the risks incurred of the solution in the large scale systems. In the thesis we adopt the hybrid approach of resource allocation for query optimization, meaning that, we first make a static resource allocation during the query compile time, and then reallocate the resources dynamically during the query runtime. As opposed to the previously proposed methods, we use a decentralized management structure. The static part of our method consists of the strategy of initial resource allocation with a query 'broker'. As for the dynamic part, we propose a strategy that uses the cooperation between relational mobile operations and stationary coordinators of nodes in order to decentralize the process of dynamic resource reallocation. Key elements of our method are: (i) limitation of research space for resolve problems caused by the large scale, (ii) principle of resources distribution between query operations for determining the parallelism degree of operations and for balancing the load dynamically and (iii) decentralization of the dynamic allocation process. Results of performance evaluation show the efficiency of our propositions. Our initial resource allocation strategy gives results superior to the referenced method that we used for the comparison. The dynamic reallocation strategy decrease considerably the response time in the presence of the system instability and load misbalance
Yesse, Thierry. "Etude du parallélisme massif appliquée à la visualisation interactive de grandes bases de données scientifiques." Rouen, 1998. http://www.theses.fr/1998ROUES086.
Full textKacimi, Rahim. "Techniques de conservation d'énergie pour les réseaux de capteurs sans fil." Thesis, Toulouse, INPT, 2009. http://www.theses.fr/2009INPT035H/document.
Full textTechnological advances during the last few years allowed the development of new and cheap sensors equiped with wireless communication which can be configured to form autonomous networks. The application areas for wireless sensor networks (WSN) are various: home automations, health care services, military domain, and environment monitoring. The imposed constraints are limited capacity of processing, storage, and especially energy. In addition, implementing WSN solutions is highly open and requires that the infrastructure, the mechanisms and the protocols should be completely designed based on each specific application.In this thesis, we first focused on small networks. We designed « Placide », a protocol stack solution for cold chain monitoring proposed within the ANR-CAPTEURS project. The first originality of this solution is based on the absence of infrastructure and base stations. « Placide » is composed of self-organizing and energy-efficient protocols based on a virtual ring construction between nodes. The second topic is devoted to an experimental study on Link Quality Indicator (LQI). There are two main objectives. Firstly, we want to endorse our precise assumptions of the first part of the work. Secondly, our poposed link quality based protocols and algorithms willbe described. We focused on LQI variations according to distance between nodes and transmission power.Thereafter, the impact of LQI on the network topology has been studied. Feedbacks are important to understand which factors affect the network lifetime. The last contribution relates to the use of these observations by proposing load balancing strategies. Our idea is that very reduced protocols and simple mechanisms can be used in routing protocols. We illustrate these principles through simple examples where we show the superiority of these solutions compared to standard routing like shortest path for example
Doghri, Inès. "Stratégies de routage multi-chemin dans les réseaux sans fil multi-sauts." Phd thesis, Ecole normale supérieure de lyon - ENS LYON, 2012. http://tel.archives-ouvertes.fr/tel-00828394.
Full textMeslmawy, Mahdi Abed Salman. "Efficient ressources management in a ditributed computer system, modeled as a dynamic complex system." Thesis, Le Havre, 2015. http://www.theses.fr/2015LEHA0007/document.
Full textGrids and clouds are types of currently widely known distributed computing systems or DCSs. DCSs are complex systems in the sense that their emergent global behavior results from decentralized interaction of its parts and is not guided directly from a central point. In our study, we present a complex system model that efficiently manages the ressources of a DCS. The entities of the DCS react to system instability and adjust their environmental condtions for optimizing system performance. The structure of the interaction networks that allow fast and reliable access to available resources is studied and improvements ar proposed
Aravinthan, Gopalasingham. "SDN based service oriented control approach for future radio access networks." Thesis, Evry, Institut national des télécommunications, 2017. http://www.theses.fr/2017TELE0013.
Full textSoftware-Defined Networking (SDN) has emerged as a new intelligent architecture for network programmability. The primary idea behind SDN is to move the control-plane outside the switches and enable external control of data-plane through a logical software entity called controller. Such approach benefits mobile network management by brining complete intelligence to the logically centralized controller. Network Function Virtualization (NFV) is the process of relocating or migrating network functions from dedicated hardware to generic servers. SDN and NFV are two closely related technologies that are often used together. The traditional mobile network architecture due to its strongest coupling between control and data planes along with limitations in scalability and flexibility requires the usage of cloud computing along with the recent revolutionary approaches in networking such as SDN and NFV to have an architecture that deploys on demand "Network-as-a-Service" for users. The global research focus of this thesis falls in to two main use cases of next generation mobile networks such as Telco and Vertical. In the telco use cases, we exploit the advantages of SDN to have flexible control framework for both Self-Organizing Networks (SON) and dynamic user processing split. In vertical use case, we apply various advantages of SDN and OpenFlow protocol to efficiently utilize the scare radio resources of wireless backhaul network in the train-to-ground communication system. Our SDN framework in general can be an efficient and alternative solution for RAN management i.e. Radio Optimization, Network Optimization, Mobility Management and Load Balancing can be achieved with such framework. Through analysis and experimentation of SDN frameworks for RAN, we shows that the proposed solutions can bring set of advantages to wireless networks such as flexibility, programmability, unified management, and enables new services
Hermann, Everton. "Simulations physiques interactives sur des architectures multi-core et multi-GPU." Grenoble, 2010. http://www.theses.fr/2010GRENM029.
Full textInteractive physical simulation is a key component of realistic virtual environments. However, the amount of computations as well as the code complexity grow quickly with the variety, number and size of the simulated objects. During this thesis we studied the different ways to improve interactivity, with the constraint of minimizing the impact in the simulation code. We started by developing a new approach to collision detection between deformable bodies that allows deep intersections. Experiments show that this approach is fast and more robust than traditional proximity-based collisions, allowing physical simulations to employ large time steps. To take profit of multi-core machines, we propose a parallelization approach that relies on a task parallelism. Prior to running a simulation step, we extract a task dependency graph that is partitioned to define the task distribution between processors. This approach has a low impact on physics algorithms as parallelism is mainly extracted from the coordination code. Finally we extended our works multi-CPU and multi-GPU architectures. Using these resources efficiently in a seamless way is a challenging issue. We propose a parallelization scheme for dynamically balancing work load between multiple CPUs and GPUs. We rely on a two level scheduling associating a traditional task graph partitioning and a work stealing guided by processor affinity and heterogeneity. These criteria are intended to limit inefficient task migrations between GPUs, the cost of memory transfers being high, and to favor mapping small tasks on CPUs and large ones on GPUs to take advantage of heterogeneity
Wion, Adrien. "Control plane in dynamic software networks." Electronic Thesis or Diss., Institut polytechnique de Paris, 2021. http://www.theses.fr/2021IPPAT007.
Full textDuring the last years, network infrastructure has moved from dedicated-hardware solutions implementing fixed functions to more flexible software based ones. On one hand, SDN (Software Defined Network) can flexibly control forwarding operations, while on the other, NFV (Network Function Virtualization) creates elastic functions that can scale with the user demands. So far, these solutions have been used to simplify network management and operations, but they let envision a network that can automatically react to network events. In this thesis, we explore to what extent these new software networks can be used to react and adapt finely to the network dynamics.Our first contribution focuses on service chaining: the ability to steer flows through a set of waypoints hosting functions before they reach their destinations. We show that a distributed control plane that relies on existing routing protocols and is constituted by autonomous nodes can dynamically steer traffic through chains of services. Our solution finely adapts its decision to the network traffic and automatically balances the induced load on the functions present in the network. Moreover, our proposal, contrary to existing solutions, can be incrementally deployed in today's network.In our second contribution, we compare two types of chaining decisions: a centralized one with an end-to-end view of the chain and a distributed approach that solely routes flow from a function to another. We show that the two decisions are close in realistic topologies. Thus, hop-by-hop chaining could be used without affecting chaining performance. Finally, we explore how software networks can react to network dynamics in datacenters. So far, load balancers use static policies to spread incoming traffic on servers, which leads to imbalance and overprovisioning. We propose to close the loop and dynamically adapt the policy to the server load variation. Our MPC (Model Predictive Control) approach proved to be efficient to reduce load imbalance at a slow pace, thus improving the number of requests a cluster can process
Kamel, Mouna. "Equilibrage des charges : application a un environnement de bases d'objets." Toulouse 3, 1991. http://www.theses.fr/1991TOU30050.
Full textGou, Changjiang. "Task Mapping and Load-balancing for Performance, Memory, Reliability and Energy." Thesis, Lyon, 2020. http://www.theses.fr/2020LYSEN047.
Full textThis thesis focuses on multi-objective optimization problems arising when running scientific applications on high performance computing platforms and streaming applications on embedded systems. These optimization problems are all proven to be NP-complete, hence our efforts are mainly on designing efficient heuristics for general cases, and proposing optimal solutions for special cases.Some scientific applications are commonly modeled as rooted trees. Due to the size of temporary data, processing such a tree may exceed the local memory capacity. A practical solution on a multiprocessor system is to partition the tree into many subtrees, and run each on a processor, which is equipped with a local memory. We studied how to partition the tree into several subtrees such that each subtree fits in local memory and the makespan is minimized, when communication costs between processors are accounted for.Then, a practical work of tree scheduling arising in parallel sparse matrix solver is examined. The objective is to minimize the factorization time by exhibiting good data locality and load balancing. The proportional mapping technique is a widely used approach to solve this resource-allocation problem. It achieves good data locality by assigning the same processors to large parts of the task tree. However, it may limit load balancing in some cases. Based on proportional mapping, a dynamic scheduling algorithm is proposed. It relaxes the data locality criterion to improve load balancing. The performance of our approach has been validated by extensive experiments with the parallel sparse matrix direct solver PaStiX.Streaming applications often appear in video and audio domains. They are characterized by a series of operations on streaming data, and a high throughput. Multi-Processor System on Chip (MPSoC) is a multi/many-core embedded system that integrates many specific cores through a high speed interconnect on a single die. Such systems are widely used for multimedia applications. Lots of MPSoCs are batteries-operated. Such a tight energy budget intrinsically calls for an efficient schedule to meet the intensive computation demands. Dynamic Voltage and Frequency Scaling (DVFS) can save energy by decreasing the frequency and voltage at the price of increasing failure rates. Another technique to reduce the energy cost and meet the reliability target consists in running multiple copies of tasks. We first model applications as linear chains and study how to minimize the energy consumption under throughput and reliability constraints, using DVFS and duplication technique on MPSoC platforms.Then, in a following study, with the same optimization goal, we model streaming applications as series-parallel graphs, which are more complex than simple chains and more realistic. The target platform has a hierarchical communication system with two levels, which is common in embedded systems and high performance computing platforms. The reliability is guaranteed through either running tasks at the maximum speed or triplication of tasks. Several efficient heuristics are proposed to tackle this NP-complete optimization problem
Kyeongja, Lee. "Modèle global pour la Qualité de Service dans les réseaux de FAI : intégration de DiffServ et de l’ingénierie de trafic basée sur MPLS." Ecole Centrale de Lille, 2006. https://pepite-depot.univ-lille.fr/LIBRE/Th_Num/2006/50376-2006-Kyeongja.pdf.
Full textMultipath routing is a mechanism for load balancing in which the total load is spatially distributed over several paths. It can reduce congestion probability given by routing into the shortest path in the current Internet. LSPs of MPLS provide path controllability by explicit route in the connectionless IP network. Multipath routing scheme is composed of two steps, the selection of multiple candidate paths and the traffic splitting among these selected paths. We compare some recent multipath routing algorithms based on MPLS on the point of scalability and stability. Among them, we pick up WDP and LDM for our hybrid approach which combines cross-steps of different algorithms. PER algorithm is proposed in the improvement of LDM’s traffic splitting step. Using LDM, WDP and PER, three hybrid QoS algorithms are made. Their performances are proved by simulation that they, especially LBWDP (Load Balancing over Widest Disjoint Paths), are more effective than LDM or MATE. For better QoS guarantee, we integrate DiffServ with our traffic engineering technique (DiffServ-aware MPLS Traffic Engineering : DS-TE). PEMS (PEriodic Multi-Step algorithm in DS-TE network) is proposed as one DS-TE model to differentiate the quality of service according to the class. For PEMS algorithm, we propose a new algorithm to select candidate paths by proper metric of each traffic class. PER algorithm distributes the traffic among its own candidate paths by dynamic metric of requested class. Simulation results using ns-2 showed that PEMS can balance the load less than LBWDP but EF and AF classes can be ensured of the better quality of service than LBWDP
Moulahi, Tarek. "Algorithmes de diffusion dans les réseaux dynamiques de capteurs sans fil." Thesis, Besançon, 2015. http://www.theses.fr/2015BESA2021/document.
Full textIn this thesis, we are interested in the task of diffusion in networks of dynamic wireless sensors RCSF. This is an essential and primordial task, since it is necessary for neighbor discovery, routing, information distribution throughout the network, node location and time synchronization.[...]
Guittenit, Christophe. "Placement d'objets multimédias sur un groupe hétérogène de dispositifs de stockage." Toulouse 3, 2002. http://www.theses.fr/2002TOU30098.
Full textThe data administration of storage system consists in providing to each application a storage space having a quality of service appropriate to the needs for this application: quality expressed in term of storage capacity, reliability and availability of storage, and of performances in access time and throughput (bandwidth). This thesis proposes to make the study of the automatic administration of a heterogeneous storage system dedicated to the service of multimedia objects. After having studied and having carried out the classification of the various policies of placement designed to exploit this type of storage system, we propose a new data placement, the EFLEX (Entrelacement FLEXible - that is "flexible interleaving") that makes it possible to jointly exploit the bandwidth and the storage capacity of the system. .
Ait, Chellouche Soraya. "Délivrance de services média suivant le contexte au sein d'environnements hétérogènes pour les réseaux médias du futur." Thesis, Bordeaux 1, 2011. http://www.theses.fr/2011BOR14415/document.
Full textUsers’ willingness to consume media services along with the compelling proliferation of mobile devices interconnected via multiple wired and wireless networking technologies place high requirements on the Future Internet. It is a common belief today that Internet should evolve towards providing end users with ubiquitous and high quality media services and this, in a scalable, reliable, efficient and interoperable way. However, enabling such a seamless media delivery raises a number of challenges. On one hand, services should be more context-aware to enable their delivery to a large and disparate computational context. On another hand, current Internet media delivery infrastructures need to scale in order to meet the continuously growing number of users while keeping quality at a satisfying level. In this context, we introduce a novel architecture, enabling a novel collaborative framework for sharing and consuming Media Services within Future Internet (FI). The introduced architecture comprises a number of environments and layers aiming to improve today’s media delivery networks and systems towards a better user experience. In this thesis, we are particulary interested in enabling context-aware multimedia services provisioning that meets on one hand, the users expectations and needs and on another hand, the exponentially growing users’ demand experienced by these services. Two major and demanding challenges are then faced in this thesis (1) the design of a context-awareness framework that allows adaptive multimedia services provisioning and, (2) the enhancement of the media delivery platform to support large-scale media services. The proposed solutions are built on the newly introduced virtual Home-Box layer in the latter proposed architecture.First, in order to achieve context-awareness, two types of frameworks are proposed based on the two main models for context representation. The markup schemes-based framework aims to achieve light weight context management to ensure performance in term of responsiveness. The second framework uses ontology and rules to model and manage context. The aim is to allow higher formality and better expressiveness and sharing. However, ontology is known to be complex and thus difficult to scale. The aim of our work is then to prove the feasibility of such a solution in the field of multimedia services provisioning when the context management is distributed among the Home-Box layer. Concerning the media services delivery enhancement, the idea is to leverage the participating and already deployed Home-Boxes disk storage and uploading capabilities to achieve service performance, scalability and reliability. Towards this, we have addressed two issues that are commonly induced by the content replication: (1) the server selection for which we have proposed a two-level anycast-based request redirection strategy that consists in a preliminary filtering based on the clients’ contexts and in a second stage provides accurate network distance information, using not only the end-to-end delay metric but also the servers’ load one and, (2) the content placement and replacement in cache for which we have designed an adaptive online popularity-based video caching strategy among the introduced HB overlay
Bineau, Yannick. "Equilibre des paiements extérieurs, intégration des marchés des capitaux et influence du change : Réflexion sur le thème de la contrainte extérieure." Lille 1, 1997. http://www.theses.fr/1997LIL12001.
Full textJabri, Issam. "Gestion dynamique des topologies sans fils." Phd thesis, Université Henri Poincaré - Nancy I, 2008. http://tel.archives-ouvertes.fr/tel-00342847.
Full textKrajecki, Michaël Gardan Yvon. "EQUILIBRE DE CHARGE DYNAMIQUE : ETUDE ET MISE EN UVRE DANS LE CADRE DES APPLICATIONS A NOMBRE FINI DE TACHES INDEPENDANTES ET IRREGULIERES /." [S.l.] : [s.n.], 1998. ftp://ftp.scd.univ-metz.fr/pub/Theses/1998/Krajecki.Michael.SMZ9818.pdf.
Full textKrajecki, Michaël. "Equilibre de charge dynamique : étude et mise en œuvre dans le cadre des applications à nombre fini de tâches indépendantes et irrégulières." Metz, 1998. http://docnum.univ-lorraine.fr/public/UPV-M/Theses/1998/Krajecki.Michael.SMZ9818.pdf.
Full textParallelism allows the use of several processors simultaneously to solve a given problem more quickly. However, to increase the power of the machine by the number of processors is an ideal which is in general very difficult to reach, because if the task allocation is not optimized, the performances of the program will be reduced. The placement of these tasks is one of the principal problems of parallelism. In this work, we study a toolkit helping the parallelization of the FIIT applications (applications with a finite number of independent and irregular tasks). The result of this toolkit is a first parallel solution which dynamically solves the problem of load balancing. The FIIT toolkit proposes various tools to the user such as : five MIMD load balancing strategies, a mathematical model based on matrix to validate the algorithms theoretically and an environment of programming on a virtual parallel machine. To illustrate the FIIT toolkit, we consider the ray tracing as case study. We showed the irregular character of the ray tracing. In particular, we highlighted that the behaviour of the application is strongly dependent on the scene to visualize. Using the FIIT environment, we parallelized it and an experimental analysis on a CM-5 was carried out. The FIIT toolkit enabled us to parallelize the ray tracing application efficiently without worrying about the target machine nor the dynamic load balancing strategy applied
Rowshannahad, Mehdi. "Qualification Management and Closed-Loop Production Planning in Semiconductor Manufacturing." Thesis, Saint-Etienne, EMSE, 2015. http://www.theses.fr/2015EMSE0784/document.
Full textIn the first part, we take a binding restriction, called qualification, present in semiconductor manufacturing as a lever for increasing flexibility and optimizing capacity utilization. A qualification determines the processing authorization of a product on a machine (like an eligibility constraint). In order to define the best qualification, the production volume should be allocated to parallel non-identical machines which are partially reconfigurable. Capacitated flexibility measures are introduced to define the best qualification which increases machine capacity utilization at most. Batching is another industrial constraint encountered in semiconductor industry. It influences workload balancing and qualification management. Several workload balancing algorithms are proposed to find the optimal workload balance of a workcenter. Variability measures are also proposed to evaluate the workload variability of a workcenter. The second part deals with closed-loop production planning. Soitec uses Smart-Cut Technology to fabricate SOI wafers. Using this technology, one of the two raw materials used to fabricate SOI wafers can be reused several times to make other SOI wafers. However, before coming back to the SOI fabrication line, the used raw material (by-product) must be reworked in another production line. An original closed-loop production planning model adapted to the supply chain specificities of Soitec is proposed, and is validated using industrial data. Based on this industrial model, a single-item uncapacitated closed-loop lot-sizing model is defined, analyzed, and a dynamic programming algorithm is proposed for a simplified version of the problem
Girardin, Eric. "Les effets de la politique budgetaire sur l'equilibre externe dans des economies interdependantes." Rennes 1, 1987. http://www.theses.fr/1987REN11014.
Full textSébastien, Nicolas. "Distribution et Parallélisation de Simulations Orientées Agents." Phd thesis, Université de la Réunion, 2010. http://tel.archives-ouvertes.fr/tel-00474213.
Full textAl, Hajj Hassan Mohamad. "Parallélisme et équilibrage de charges dans le traitement de la jointure sur des architectures distribuées." Phd thesis, Université d'Orléans, 2009. http://tel.archives-ouvertes.fr/tel-00465073.
Full textLebel, Guilhem. "Prévision des charges aéromécaniques des rotors d'hélicoptère : Application aux pales à double flèche." Thesis, Lyon, INSA, 2012. http://www.theses.fr/2012ISAL0025.
Full textNew generation blades have led to new load computation problems due to the evolution of the general shape, with forward and backward sweep. The BlueEdgeTM blade pattented by Eurocopter imposes to reconsider the development methodology and thus it is no longer possible to speak of straight blades and the models used for load computation have to be evaluated. The objective of this thesis is to determine what has to be modified and improved in current load computation methodology in order to reach an acceptable predictive level. This work considers both aerodynamic and dynamic models implemented in the HOST multi-body computer code. The aerodynamics models are based on the hypothesis of a two dimensional flow. The use of the CFD software \emph{elsA} is evaluated. Attention is given to rotor dynamics models that have an impact on loads, such as lead-lag damper models, blade element models and hub models. This thesis presents the different models and gives orientations relating to efficient load computation methodology. The aerodynamics models are compared to windtunnels experiments from the literature. This study leads also to perform flight tests and to investigate the dampers behavior on test benches in order to confront the computed loads to the reality of the helicopter operation. The proposed methodology is able to compute with a good accuracy rotor loads for stabilized flight cases
Zouabi, Oussama. "L'impact du changement climatique la production agricole et la croissance économique : Cas de la Tunisie." Thesis, Toulon, 2015. http://www.theses.fr/2015TOUL2008.
Full textThe aim of this research is to analysis both direct and indirect effect of the climate change on the agricultural production and the economic growth. This PhD research we will be based on an empirical methodology, through applying a spacial econometrics and the cointegration technique of a panel data as well as a multisectoral general equilibrium growth model. The first part is devoted to find the effect of the climate change on the agricultural production in a macro spatial level during the period 1980-2012. The main gaol of this first chapter of this PhD is to determine the direct and indirect effect of the weather forecast and the temperature changes in one region compared to the neighbouring regions. The originality of this spacial modelisation is to give an effective spacial effect. The second part of this research is aimed to use a micro spacial analysis to find out the effect of the climate change on the agricultural production in the long term way and for each region rather then all of them together as in the first chapter. In the last part of this PhD research we developed a general and dynamic equilibrium model for the Tunisian economy taking in a count eventual climate change forcast from 2015 to 2030. Aiming to calculate the effect of this change on the agricultural production and the economic aggregate in Tunisa. The resulats show that the climate change forecast in the futur will affect the agricultural production in Tunisia in both the short run and the long term and this is because of the irrigated and non irrigated plantations. The decrease of the production will affect the food industry which will decrease the level of the investment, the employment in this sector as well as the consumption
García, Martinez Marta. "Development and validation of the Euler-Lagrange formulation on a parallel and unstructured solver for large-eddy simulation." Thesis, Toulouse, INPT, 2009. http://www.theses.fr/2009INPT006H/document.
Full textParticle-laden flows occur in industrial applications ranging from droplets in gas turbines tofluidized bed in chemical industry. Prediction of the dispersed phase properties such as concentration and dynamics are crucial for the design of more efficient devices that meet the new pollutant regulations of the European community. The objective of this thesis is to develop an Euler-Lagrange formulation on a parallel and unstructured solver for large- eddy simulation. This work is motivated by the rapid increase in computing power which opens a new way for simulations that were prohibitive one decade ago. Special attention is taken to keep data structure simplicity and code portability. Developments are validated in two configurations : an academic test of a decaying homogeneous isotropic turbulence and a polydisperse two-phase flow of a confined bluff body. The use of load-balancing capabilities is highlighted as a promising solut! ion in Lagrangian two-phase flow simulations to improve performance when strong imbalance of the dispersed phase is present
Jmila, Houda. "Dynamic resource allocation and management in virtual networks and Clouds." Thesis, Evry, Institut national des télécommunications, 2015. http://www.theses.fr/2015TELE0023/document.
Full textCloud computing is a promising technology enabling IT resources reservation and utilization on a pay-as-you-go manner. In addition to the traditional computing resources, cloud tenants expect compete networking of their dedicated resources to easily deploy network functions and services. They need to manage an entire Virtual Network (VN) or infrastructure. Thus, Cloud providers should deploy dynamic and adaptive resource provisioning solutions to allocate virtual networks that reflect the time-varying needs of Cloud-hosted applications. Prior work on virtual network resource provisioning only focused on the problem of mapping the virtual nodes and links composing a virtual network request to the substrate network nodes and paths, known as the Virtual network embedding (VNE) problem. Little attention was paid to the resource management of the allocated resources to continuously meet the varying demands of embedded virtual networks and to ensure efficient substrate resource utilization. The aim of this thesis is to enable dynamic and preventive virtual network resources provisioning to deal with demand fluctuation during the virtual network lifetime, and to enhance the substrate resources usage. To reach these goals, the thesis proposes adaptive resource allocation algorithms for evolving virtual network requests. We adress the extension of an embedded virtual node requiring more resources and consider the substrate network profitability. We also deal with the bandwidth demand variation in embedded virtual links. We first provide a heuristic algorithm to deal with virtual nodes demand fluctuation. The work is extended by designing a preventive re-configuration scheme to enhance substrate network profitability. Finally, a distributed, local-view and parallel framework was devised to handle embedded virtual links bandwidth fluctuations. The approach is composed of a controller and three algorithms running in each substrate node in a distributed and parallel manner. The framework is based on the self-stabilization approach, and can manage various forms of bandwidth demand variations simultaneously
Ramos, Sara Isabel Pulga. "Os planos para a igualdade nas empresas do setor empresarial do estado : do cumprimento legal ao potencial transformador." Master's thesis, Instituto Superior de Economia e Gestão, 2020. http://hdl.handle.net/10400.5/20772.
Full textApesar dos alicerces legais existentes em matéria de igualdade de género, as assimetrias entre mulheres e homens, em particular nas organizações, tendem a subsistir. A literatura sobre género e mudança organizacional tem defendido a integração de uma perspetiva de género, através de um diagnóstico interno e da adoção de uma estratégia para a igualdade consubstanciada plano de ação de modo a reduzir os efeitos dessas assimetrias. Para que a mudança seja efetiva, os planos devem permitir alcançar uma efetiva igualdade entre mulheres e homens nas respetivas organizações. O presente estudo foi impulsionado pelo atual quadro normativo, que determina que as empresas cotadas em bolsa e as empresas do setor público empresarial (setor empresarial do Estado e setor empresarial local) estão obrigadas a divulgar anualmente os seus planos para a igualdade. A investigação realizada incide exclusivamente sobre o setor empresarial do Estado e propõe-se verificar se as empresas cumprem com este requisito legal e se os planos divulgados se revestem de potencial de mudança. Pretende-se ainda relacionar a divulgação desses documentos com o nível de equilíbrio entre mulheres e homens nos órgãos de governo das empresas do segmento em estudo. Conclui-se que a grande maioria das empresas não cumpre com os dispostos legais em vigor, além de a generalidade dos planos se revestir de um potencial transformador reduzido a moderado. Concluímos ainda não existir relação entre a sua elaboração/divulgação e o nível de equilíbrio entre mulheres e homens nos órgãos de governo das empresas.
Despite the current legal foundation in the field of gender equality, the gap between women and men within organizations seems to persist. The literature on gender and organizational change have been advocating the presence of a gender perspective which should begin by an internal diagnosis followed by the adoption of an equality driven strategy which is to be backed by an action plan. To achieve the effective equality between women and men, this plan needs to be able to enable actual change. The present study was driven by the current legal framework, which dictates that all listed companies and public companies in the business sector (state and local) are to share their plans for equality on an annual basis. The research was solely carried out on the state's business sector, and it intended to ascertain if companies are following these legal requirements, as well as to assess the potential for change of these strategic documents. It was also our goal to check for a correlation between the development and disclosure of these strategic documents and the gender balance within the governance bodies of these companies. We have come to the conclusion that the vast majority of companies do not comply with the legal requirements, and that most of the plans for equality have a moderate to low potential for change. In addition, it is possible to confirm that there is no correlation between the creation/disclosure of the plan and the gender balance in the governance bodies of the researched companies.
info:eu-repo/semantics/publishedVersion
Tacheix, Thierry. "L'influence des variables financières sur l'emploi : France 1970-1988." Clermont-Ferrand 1, 1991. http://www.theses.fr/1991CLF10007.
Full textRamdane, Yassine. "Big Data Warehouse : model de distribution des cubes de données à la volée." Thesis, Lyon, 2019. http://www.theses.fr/2019LYSE2099.
Full textPartitioning and distributing of the data have been widely used in sharing nothing systems, more particularly in the distributed systems that used the MapReduce paradigm, such as Hadoop ecosystem and Spark. They have been used for many purposes, such as load balancing, skipping to load unnecessary data partitions and for guiding the physical design ofdistributed databases or data warehouses. To do analysis with data warehouses, usually, we used OLAP queries. An OLAP query is a complex query that contains several cost operations, such as the star join, the projection, filtering, and aggregate functions. In this thesis, we propose different static and dynamic approaches of partitioning and load balancing of the data, to improve the performances of distributed big data warehouses over Hadoop cluster. We have proposed different static and dynamic schemes of a big data warehouse over a cluster of homogeneous nodes, which can help the distributed system to enhance the executing time of OLAP query operations, such as star join operation, scanning tables, and Group-By operation.We have proposed four approaches: The first approach, is a new data placement strategy which able to help a query processing system to perform a star join operation in only one MapReduce cycle, without a shuffle phase; In the second contribution, we propose different partitioning and bucketing techniques to skip loading some HDFS blocks and to enhance the parallel treatment of the distributed system, based on a workload-driven model; In the third approach, we propose a novel physical design of distributed big data warehouse over Hadoop cluster, such as we combine between our first data-driven approach and the second workloaddriven solution; The fourth contribution has been developed to improve Group-by and aggregate functions, by using a dynamic method, which able to define on the fly the best partitioning scheme of the reducer inputs. To evaluate our approaches, we have conducted some experiments on different cluster sizes, using different data warehouses volumes where the fact table has more than 28 billions of records. We have used the TPC-DS benchmark, a Hadoop-YARN platform, a Spark engine, and Ray and Hive system. Our experiments show that our methods outperform the state-of-the-art approaches in many aspects, especially on the OLAP query execution time
Prelle, Thomas. "Gestion optimisée d'un modèle d'agrégation de flexibilités diffuses." Thesis, Nantes, Ecole des Mines, 2014. http://www.theses.fr/2014EMNA0177/document.
Full textThe desire to increase the share of renewable energies in the energy mix leads to an increase inshare of volatile and non-controllable energy and makes it difficult to meet the supply-demand balance. A solution to manage anyway theses energies in the current electrical grid is to deploy new energy storage and demand response systems across the country to counter balance under or over production. In order to integrate all these energies systems to the supply and demand balance process, there are gathered together within a virtual flexibility aggregation power plant which is then seen as a virtual power plant. As for any other power plant, it is necessary to compute its production plan. Firstly, we propose in this PhD thesis an architecture and management method for an aggregation power plant composed of any type of energies systems. Then, we propose algorithms to compute the production plan of any types of energy systems satisfying all theirs constraints. Finally, we propose an approach to compute the production plan of the aggregation power plant in order to maximize its financial profit while complying with all the constraints of the grid
Steffen, Jean-Sébastien. "Modélisation tridimensionnelle globale du squelette pour l'aide au diagnostic et à la prise en charge thérapeutique des pathologies rachidiennes affectant l'équilibre postural." Phd thesis, 2011. http://pastel.archives-ouvertes.fr/pastel-00649423.
Full text