Dissertations / Theses on the topic 'Réseaux de stockage (Informatique)'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Réseaux de stockage (Informatique).'
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.
Modrzejewski, Remigiusz. "Distribution et Stockage de Contenus dans les Réseaux." Phd thesis, Université Nice Sophia Antipolis, 2013. http://tel.archives-ouvertes.fr/tel-00917032.
Full textLa, Chi Anh. "Réplication de contenu dans les réseaux sans fil mobiles." Paris, Télécom ParisTech, 2010. http://pastel.archives-ouvertes.fr/pastel-00545009.
Full textThe growth of mobile devices and network-based services nowadays has raised a timely question on how to efficiently distribute the data items to mobile users. Network applications need data as an input to process and provide information to users. Consequently, data traffic exerted by mobile devices fetching content is a drainage of mobile operators’ network resources. Mobile users are now coping with the congestion at network gateways and due to the unpredictability of human mobility, mobile service providers cannot sufficiently provision infrastructures for their customers. Content replication in this context has been proved as a good solution to enhance network performance and scalability. In this thesis, we tackle the issues of content replication in heterogeneous mobile networks. Such scheme requires us to solve two basic questions: where and how many replicas should be placed in the system. We study the solution through the lenses of facility location theory and design a distributed mechanism that reduces content access latency and avoids congestion at mobile gateways. Additionally, we consider the resource constraints of mobile devices and introduce a P2P cache-and-forward mechanism for load balancing purpose. We evaluate our mechanisms against realistic human mobility models. Finally, to address rational users who may behave selfishly in replicating content, we derive a cost model and study content replication scheme using tools akin to game theory. Based on the theoretical findings, our future work is to develop the strategies to be implemented in a practical network setting
Baud, Loïc. "ROSA, un réseau de recouvrement adaptable, extensible et auto-organisant." Paris, Télécom ParisTech, 2010. https://pastel.hal.science/pastel-00006101.
Full textThe works presented in this thesis take place within the context of the overlay network. An overlay network is a network built on top another network. To each node of the overlay network corresponds a node of the underlying network. These overlay networks have been popularized with the rise of the peer-to-peer networks (that belong to a sub-class of the overlay networks) in the years 2000. There exists many kinds of overlay networks, some are scalable, some ensure a resilient topology, some offers a reliable routing service, etc. But none of the existing overlay network are adaptable to different kind of covered network and propose a large set of services. Any existing overlay network is only dedicated to a particular task. One can imagine an overlay network that could be deployed astride many different kind of underlying network and not dedicating to only a particular task. My thesis aims to define the basis and develop such an overlay network
Romito, Benoit. "Stockage décentralisé adaptatif : autonomie et mobilité des données dans les réseaux pair-à-pair." Caen, 2012. http://www.theses.fr/2012CAEN2072.
Full textWan, Kai. "Limites fondamentales de stockage pour les réseaux de diffusion de liens partagés et les réseaux de combinaison." Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLS217/document.
Full textIn this thesis, we investigated the coded caching problem by building the connection between coded caching with uncoded placement and index coding, and leveraging the index coding results to characterize the fundamental limits of coded caching problem. We mainly analysed the caching problem in shared-link broadcast model and in combination networks. In the first part of this thesis, for cache-aided shared-link broadcast networks, we considered the constraint that content is placed uncoded within the caches. When the cache contents are uncoded and the user demands are revealed, the caching problem can be connected to an index coding problem. We derived fundamental limits for the caching problem by using tools for the index coding problem. A novel index coding achievable scheme was first derived based on distributed source coding. This inner bound was proved to be strictly better than the widely used “composite (index) coding” inner bound by leveraging the ignored correlation among composites and the non-unique decoding. For the centralized caching problem, an outer bound under the constraint of uncoded cache placement is proposed based on the “acyclic index coding outer bound”. This outer bound is proved to be achieved by the cMAN scheme when the number of files is not less than the number of users, and by the proposed novel index coding achievable scheme otherwise. For the decentralized caching problem, this thesis proposes an outer bound under the constraint that each user stores bits uniformly and independently at random. This outer bound is achieved by dMAN when the number of files is not less than the number of users, and by our proposed novel index coding inner bound otherwise. In the second part of this thesis, we considered the centralized caching problem in two-hop relay networks, where the server communicates with cache-aided users through some intermediate relays. Because of the hardness of analysis on the general networks, we mainly considered a well-known symmetric relay networks, combination networks, including H relays and binom{H}{r} users where each user is connected to a different r-subset of relays. We aimed to minimize the max link-load for the worst cases. We derived outer and inner bounds in this thesis. For the outer bound, the straightforward way is that each time we consider a cut of x relays and the total load transmitted to these x relays could be outer bounded by the outer bound for the shared-link model including binom{x}{r} users. We used this strategy to extend the outer bounds for the shared-link model and the acyclic index coding outer bound to combination networks. In this thesis, we also tightened the extended acyclic index coding outer bound in combination networks by further leveraging the network topology and joint entropy of the various random variables. For the achievable schemes, there are two approaches, separation and non-separation. In the separation approach, we use cMAN cache placement and multicast message generation independent of the network topology. We then deliver cMAN multicast messages based on the network topology. In the non-separation approach, we design the placement and/or the multicast messages on the network topology. We proposed four delivery schemes on separation approach. On non-separation approach, firstly for any uncoded cache placement, we proposed a delivery scheme by generating multicast messages on network topology. Moreover, we also extended our results to more general models, such as combination networks with cache-aided relays and users, and caching systems in more general relay networks. Optimality results were given under some constraints and numerical evaluations showed that our proposed schemes outperform the state-of-the-art
Pamba, Capo-Chichi Medetonhan Shambhalla Eugène William. "Conception d’une architecture hiérarchique de réseau de capteurs pour le stockage et la compression de données." Besançon, 2010. http://www.theses.fr/2010BESA2031.
Full textRecent advances in various aeras related to micro-electronics, computer science and wireless networks have resulted in the development of new research topics. Sensor networks are one of them. The particularity of this new research direction is the reduced performances of nodes in terms of computation, memory and energy. The purpose of this thesis is the definition of a new hierarchical architecture of sensor networks usable in different contexts by taking into account the sensors constraints and providing a high quality data such as multimedia to the end-users. We present our hierachical architecture with different nodes and the wireless technologies that connect them. Because of the high consumtpionof data transmission, we have developped two data compression algortithms in order to optimize the use of the channel by reducing data transmitted. We also present a solution for storing large amount of data on nodes by integrating the file system FAT16 under TinyOS-2. X
Ribeiro, Heverson Borba. "Exploiting rateless coding in structured overlays to achieve persistent storage." Rennes 1, 2012. http://www.theses.fr/2012REN1S166.
Full textThe substantial increase in the amount of information over the Internet has contributed to an extraordinary demand for persistent data storage. Centralized storage architectures are expensive, weakly scalable and vulnerable to attacks as they represent single points of failure in the system. Over last few years, peer-to-peer architectures have emerged as an alternative for implementing persistent data-storage. Open peer-to-peer systems are fundamentally scalable and cheaper than client-server approaches. However, in order to successfully build persistent storage systems using the peer-to-peer approach two fundamental challenges need to be addressed. A) To cope with the transient connectivity of peers. B) To reduce the impact of misbehaving peers. Replication is a common approach used to cope with transient connectivity in peer-to-peer storage systems. However, depending on the frequency peers join and leave the system this approach can present negative impacts in terms of storage overhead and bandwidth consumption. Peer-to-peer overlays that focus on tolerating the presence of Byzantine peers usually make the assumption that no more than a bounded fraction of peers in the system are malicious. However, estimating the proportion of malicious peers in open peer-to-peer system is not reliable. Thus, finding a scalable architecture to provide reliable and persistent data storage while coping with these issues is aninteresting achievement. In this thesis we present the design of Datacube. Datacube is an efficient and scalable peer-to-peer storage architecture that provides data persistence by implementing a hybrid redundancy scheme on top of a cluster-based structured overlay. The hybrid redundancy scheme proposed by Datacube ensures data persistence and integrity despite the intermittent connection of peers and the presence of adversarial peers. Datacube relies on the properties of the new class of rateless erasure codes to implement its hybrid redundancy scheme. The analytical evaluations have shown that Datacube performs notably well in terms of availability, storage overhead and bandwidth. Additionally, empirical evaluations have shown the performance of rateless erasure codes in the context of peer-to-peer storage systems. These evaluations helped to understand how the coding parameters impact on the performance of the architecture. To the best of our nowledge, this is the first comprehensive study that helps application designers in finding the values for the coding parameters to best fit their peer-to-peer context
Cutillo, Leucio Antonio. "Protection des données privées dans les réseaux sociaux." Electronic Thesis or Diss., Paris, ENST, 2012. http://www.theses.fr/2012ENST0020.
Full textOnline Social Network (OSN) applications allow users of all ages and educational background to easily share a wide range of personal information with a theoretically unlimited number of partners. This advantage comes at the cost of increased security and privacy exposures for users, since in all existing OSN applications, to underpin a promising business model, users' data is collected and stored permanently at the databases of the service provider, which potentially becomes a “Big Brother” capable of exploiting this data in many ways that can violate the privacy of individual users or user groups. This thesis suggests and validates a new approach to tackle these security and privacy problems. In order to ensure users' privacy in the face of potential privacy violations by the provider, the suggested approach adopts a distributed architecture relying on cooperation among a number of independent parties that are also the users of the online social network application. The second strong point of the suggested approach is to capitalize on the trust relationships that are part of social networks in real life in order to cope with the problem of building trusted and privacy-preserving mechanisms as part of the online application. Based on these main design principles, a new distributed Online Social Network, namely Safebook, is proposed: Safebook leverages on real life trust and allows users to maintain the control on the access and the usage of their own data. The prototype of Safebook is available at www.safebook.eu
Cutillo, Leucio Antonio. "Protection des données privées dans les réseaux sociaux." Phd thesis, Télécom ParisTech, 2012. http://pastel.archives-ouvertes.fr/pastel-00932360.
Full textMechraoui, Salah-Eddine. "Étude et développement du contrôle par émission acoustique des phénomènes de corrosions et de fuites appliqué au stockage des produits pétroliers : cas des bacs de stockage." Compiègne, 2007. http://www.theses.fr/2007COMP1728.
Full textLn the petrochemical industry the storage tanks under operation are very often subjected to high mechanical stress, fatigue, aging and corrosion phenomena reducing their lifespan. These phenomena deteriorate these structures severely and lead generally to catastrophic leaks for nature and significant economic lasses. Therefore, the question of security and safety of such installations became a real priority and the monitoring with different non-destructive testing techniques (NDT) is necessity, such as the acoustic emission (AE). This innovative technique has major advantages in the detection of leaks and corrosion in storage tanks. It’s a global technique which can be applied on line without significant immobilization of the structure. However, to succeed a correct diagnosis, the mastery of this technique is very important for a good results interpretation. Lt is in this way that this thesis was realised. The main objective is to study and analyse separately, in the laboratory scale, the different phenomena sources of the AE in storage tanks such as : new corrosion, old corrosion and leaks. The new corrosion tests are realised on low carbon steel S235JR used in real crude oil manufacturing. The old corrosion tests have been carried out on real steel specimen cut from a real storage tank. Finally leaks are simulated on a S235JR mini-structure. Two type of data classification are used for the treatment : classification by a neural networks classification and a pattern recognition classification of AE events collected from laboratory tests. Based on these results, an approach to the industrial scale problem is proposed, applied and analysed
Khelil, Amar. "Elaboration d'un système de stockage et exploitation de données pluviométriques." Lyon, INSA, 1985. http://www.theses.fr/1985ISAL0034.
Full textThe Lyon District Urban Area (CO. UR. LY. ) may be explained from an hydrological point of view as a 600 km2 area equipped with a sewerage system estimated by 2 000 km of pipes. Due to the complexity of the sewerage network of the area, it must therefore be controlled by an accurate and reliable system of calculation to avoid any negative consequences of its function. The capacity of the present computerising system SERAIL, allows an overall simulation of the functioning of drainage / sewerage system. This model requires an accurate information of the rainfall rate which was not previously available. Therefore a 30 rain gages network (with cassette in sit recording) was set up within the Urban District Area in 1983. This research however introduces the experiment of three steps: 1) to install the network; 2) to build up a data checking and storage system; 3) to analyse the data. The characteristic nature of this work deals with the data analysis system. It allows to extract easily and analyse any rainfall event important to the hydrologist. Two aims were defined: 1) to get a better understanding of the phenomena (punctual representations ); 2) to build up models. In order to achieve the second aim, it was necessary to think about the fitting of the propounded models and their limits which led to the setting up of several other programmes for checking and comparison. For example a complete analysis of a rainfall event is given with comments and conclusion
Chikhaoui, Amina. "Vers une approche intelligente de placement de données dans un cloud distribué basé sur un système de stockage hybride." Electronic Thesis or Diss., Brest, 2022. http://www.theses.fr/2022BRES0024.
Full textCloud federation makes it possible to seamlessly extend the resources of Cloud Service Providers (CSP) in order to provide a better Quality of Service (QoS) to customers without additional deployment costs. Storage as a Service (StaaS), is one of the main Cloud services offered to customers. For such a service, storage Input/Output (I/O) performance and network latency are among the most important metrics considered by customers. In effect, transactions for some database queries spend 90% of the execution time in I/O operations. In order to satisfy customers, some Cloud companies already include latency guarantees in their Service Level Agreements (SLA) and customers can pay additional fees to further reduce latency. This thesis addresses the data placement problem for a CSP that is part of a federation. Indeed,offering attractive and inexpensive services is a big challenge for CSP. Our goal is to pro-vide intelligent approaches for a better data placement that minimizes the cost of placement for the provider while satisfying the customers QoS requirements.This approach must take into account the heterogeneity of internal and external storage resources in terms of several parameters (such as capacity, performance, pricing) as well as customer characteristics and requirements.Despite the fact that many data placement strategies have been proposed for hybrid storage systems, they are not generalizable to every architecture. Indeed, a placement strategy must be designed according to the system architecture for which it is proposed and the target objectives
Tabart, Quentin. "Modélisation et commande d'une interface à base d'un convertisseur NPC pour système de stockage hybride dédié micro-réseau." Thesis, Université Grenoble Alpes (ComUE), 2019. http://www.theses.fr/2019GREAT076.
Full textThe increasing rise of distributed generation integration in the energy mix is a challenging issue. Indeed distributed generation based on renewable energy sources like solar and wind creates fluctuation of the grid’s parameter and can lead to minor issues like harmonics, over or under voltages, flicker and so on, or to major failure like blackouts. The concept of microgrid associated with energy storage systems is seen as a potential solution to increase renewable energy integration to the grid, while increasing efficiency and reliability at the same time.A review of energy Storage technologies is carried out and it results that currently none of these technologies can offer energy and power capabilities while fast response and efficiency at the same time. The use of Hybrid Energy Storage System allows to reach such performance. For this thesis, a Li-Ion and a Vanadium Redox Flow Battery have been selected for their good power and energy abilities respectively, and their high round trip efficiency.The main contribution of this work lies in the structural limits analysis of the topology and the control algorithm for the four leg three level NPC used as a multipurpose interface between a Renewable Energy Source, a Hybrid Energy Storage System and a Microgrid. The effectiveness of the proposed algorithms and limits model are tested both in simulation and experiments. Ancillary services like frequency and voltage support are then developed and tested in simulation under various conditions
Durou, Hugo. "Vers l'autonomie énergétique des réseaux de capteurs embarqués : conception et intégration d'un générateur piézoélectrique et d'un micro dispositif de stockage capacitif en technologie silicium." Phd thesis, Toulouse 3, 2010. http://thesesups.ups-tlse.fr/2803/.
Full textWireless sensor networks offer very interesting possibilities for structural health monitoring, especially on aircrafts. However the sensors in these networks only have limited onboard energy resources, which is a severe restriction to their autonomy over a wide lifespan. As electronics and sensors power consumption reduces, a possible - and explored by many research teams for the last decade - solution is to harvest energy from the immediate environment of the microsystem, to store it and use it when needed to supply the sensor and electronics. In this thesis we propose to harvest energy from engine vibrations on an aircraft structure to supply a structure health monitoring sensor. Our contribution lies in the conception and integration on Silicon of a miniature piezoelectric harvester and capacitive charge storage device. As for the piezoelectric harvester, a proposed finite elements analysis coupled with a SPICE description of the charge circuit enabled the design of an optimized device, which is made of 4 monomorphic cantilevers (Si/PZT) that can generate > µW power and > V voltage despite low power incoming vibrations (0. 1 to 0. 5g @ 40-80 Hz). This device has been fabricated on a silicon wafer using standard MEMS technologies and femtosecond LASER etching of the cantilevers. The charge storage device, designed and integrated onto silicon, is an electrochemical double layer capacitor. Development of the geometry of the electrodes, the active material deposition and hermetic wafer level sealing under water-free atmosphere of the component have been led. A VHDL-AMS model of both micro-devices (harvester and charge storage) is proposed and systems simulations over simple use cases are compared to the experiment
Durou, Hugo. "Vers l'autonomie énergétique des réseaux de capteurs embarqués : conception et intégration d'un générateur piézoélectrique et d'un micro dispositif de stockage capacitif en technologie silicium." Phd thesis, Université Paul Sabatier - Toulouse III, 2010. http://tel.archives-ouvertes.fr/tel-00578741.
Full textKhan, Junaid. "Large scale data collection and storage using smart vehicles : An information-centric approach." Thesis, Paris Est, 2016. http://www.theses.fr/2016PESC1045/document.
Full textThe growth in the number of mobile devices today result in an increasing demand for large amount of rich multimedia content to support numerous applications. It is however challenging for the current cellular networks to deal with such increasing demand, both in terms of cost and bandwidth for the ``massive'' content generated and consumed by mobile users in an urban environment due to its connection-centric nature. The technological advancement in modern vehicles allow us to harness their computing, caching and communication capabilities to supplement infrastructure network. It is now possible to recruit smart vehicles to collect, store and share heterogeneous data on urban streets in order to provide citizens with different services. Therefore, we leverage the recent shift towards Information Centric Networking (ICN) to introduce two schemes, VISIT and SAVING for the efficient collection and storage of content at vehicles, closer to the urban mobile user to avoid bandwidth and cost. VISIT is a platform which defines novel centrality metrics based on the social interest of urban users to identify and select the appropriate set of best candidate vehicles to perform urban data collection. SAVING is a social-aware data storage system which exploits complex networks to present game-theoretic solutions for finding and recruiting vehicles adequate to perform collaborative content caching in an urban environment. VISIT and SAVING are simulated for around 2986 vehicles with realistic urban mobility traces and comparison results with other schemes in literature suggest both not only efficient but also scalable data collection and storage systems
Ortiz, Aurélien. "Contrôle de la concurrence dans les grilles informatiques." Toulouse 3, 2009. http://thesesups.ups-tlse.fr/770/.
Full textThese last decades, the progress made in the field of telecommunications made possible the gathering of multiple clusters, interconnected by a wide area network. The birth of grid computing allowed the collaboration of geographically distributed resources, thus authorizing the execution of applications which require a great power of calculation and a large storage space. A middleware is then used to federate the grid resources and to solve the problems involved in the heterogeneity and dynamicity of those resources, as well as the safety of the grid's administrative domains. The ViSaGe project was born in this context. ViSaGe is a grid middleware which includes a distributed file system, and a virtualization layer which aggregates the storage resources dispersed among the grid nodes, in order to provide the user a global view of a huge shared virtual space. In this thesis, we were interested in the ViSaGe's concurrency control service : the VCCC. This component uses a mutual exclusion algorithm to ensure consistency accesses to various ressources shared by the grid nodes. This service is essential, but it generates a lot of control messages on the network. However, these messages, which are often of small size, are strongly exposed to the latency of the network which characterizes grid environment. Consequently, the concurrency control very often degrades the performance and the reactivity of the middleware. What we achieved in this thesis consists in bringing a competitive solution to carry out the synchronization of the grid nodes. We proposed a mutual exclusion algorithm based on several techniques resulting from the literature. In particular, we used a token algorithm, for which the grid nodes are organized according to a tree structure. Moreover, we implemented other methods to make easier the adaptation of the VCCC to the architecture of the grid, and thus to enhance the performance of our middleware. First of all, we improved the cache management of the other components of ViSaGe, according to the workload observed in the VCCC. In addition, we optimized the way that the concurrency control is spread out over the various grid nodes. Lastly, we developed a method based on the use of the multicast, which makes it possible in some cases to go round the problem induced by the network latency, with an aim of improving the reactivity of the system. .
Souza, Oliveira Camila Helena. "Reliability and cost efficiency in coding-based in-network data storage and data retrieval for IoT/WSNs." Thesis, Paris Est, 2015. http://www.theses.fr/2015PESC1134/document.
Full textWireless Sensor Networks (WSN) are made up of small devices limited in terms of memory, processing and energy capacity. They work interconnected and autonomously in order to monitoring a region or an object of interest. The evolution in the development of devices more powerful (with new capability such as energy harvesting and acting) and less expensive made the WSNs a crucial element in the emergence of Internet of Things (IoT). Nonetheless, assuming the new applications and services offered in the IoT scenario, new issues arise in the data management performed in the WSNs. Indeed, in this new context, WSNs have to deal with a large amount of data, now consumed on-demand, while ensure a good trade-off between its reliability and retrievability, and the energy consumption. In the scope of this thesis, we are interested in the data management in the WSN in the context of IoT realm. Specifically, we approach the problem of in-network data storage by posing the following question: How to store data for a short term in the WSNs so that the data could be easily retrieved by the consumers while ensuring the best trade-off between data reliability and conservation of energy resources? Foremost, we propose a reliable data storage scheme based on coding network, and assuming a communication model defined by the Publish/Subscribe paradigm. We validate the efficiency of our proposal by a theoretical analyses that is corroborate by a simulation evaluation. The results show that our scheme achieves a reliability of 80% in data delivery with the best cost-benefit compared to other data storage scheme. Aiming to further improve the performance of the data storage scheme proposed in our first contribution, we propose its optimization (modeling it as a Markov Decision Process (MDP)) in order to store data with optimal trade-off between reliability and communication overhead (in this context, also seen as energy consumption), and in an autonomously and adaptive way. For the best of our knowledge, our optimized data storage scheme is the only to ensure data reliability while adapt itself according to the service requirements and network condition. In addition, we propose a generalization of the mathematical model used in our first contribution, and a system model that defines the integration of WSNs performing our data storage scheme in the context for which it was envisaged, the IoT realm. Our performance evaluation shows that our optimization allows the consumers to retrieve up to 70% more packets than a scheme without optimization whereas increase the network lifetime of 43%.Finally, after being interested in finding the best trade-off between reliability and cost, we now focus on an auxiliary way to reduce the energy consumption in the sensor nodes. As our third contribution, we propose a study, in two parts, to measure how much a node activity scheduling can save energy. First, we propose an improvement in the duty cycle mechanism defined in the 802.15.4. Then, we propose a duty cycle mechanism introduced into our data storage scheme aiming at saving energy in the storage nodes. The simulation results show that our solution to the duty cycle mechanism in 802.15.4 led in considerable saving in energy costs. However, regarding duty cycle in our data storage scheme, it did not end up in more energy saving. Actually, as our optimized scheme already saves as much resource energy as possible while ensuring high reliability, the duty cycle mechanism can not improve the energy saving without compromise the data reliability. Nonetheless, this result corroborates that our scheme, indeed, performs under the optimal trade-off between reliability and communication overhead (consumption energy)
Ruty, Guillaume. "Towards more scalability and flexibility for distributed storage systems." Electronic Thesis or Diss., Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLT006.
Full textThe exponentially growing demand for storage puts a huge stress on traditionnal distributed storage systems. While storage devices' performance have caught up with network devices in the last decade, their capacity do not grow as fast as the rate of data growth, especially with the rise of cloud big data applications. Furthermore, the performance balance between storage, network and compute devices has shifted and the assumptions that are the foundation for most distributed storage systems are not true anymore. This dissertation explains how several aspects of such storage systems can be modified and rethought to make a more efficient use of the resource at their disposal. It presents an original architecture that uses a distributed layer of metadata to provide flexible and scalable object-level storage, then proposes a scheduling algorithm improving how a generic storage system handles concurrent requests. Finally, it describes how to improve legacy filesystem-level caching for erasure-code-based distributed storage systems, before presenting a few other contributions made in the context of short research projects
Deniel, Philippe. "NFS-Ganesha ˸ évolutions d'un serveur NFS pour le HPC du Terascale à l'Exascale." Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASG056.
Full textThis thesis exposes NFS-Ganesha, a NFS server running in userland, dedicated toHPC, as well as its evolutions from the Terascale period of the early 2000 years to Today'sExascale periode.Born because of operational needs, with close relations to the exploitation of large HPCstorage systems, NFS-Ganesha was designed to be generic and parallel. The apparition of large parallel files systels, giving birth to « data centric » architectures, and the joint apparition of the NFSv4 protocol made NFS-Gaesha a generic NFS server capable of being iterfaced to many different backends. The evoluion of NFSv4, via NFSv4.1 and its new pNFS feature made NFS-Ganesha a tandard adopted by a strong open-source community involving researchers and industrials.NFS-Ganesha will be used to build the IOProxy feature, with the creation of new parallelprotocols. Involved in different european R&D projects. NFS-Ganesha will be used toimplement epehemeral services to answer the requirements of Exascale systems
Pawlowski, Filip igor. "High-performance dense tensor and sparse matrix kernels for machine learning." Thesis, Lyon, 2020. http://www.theses.fr/2020LYSEN081.
Full textIn this thesis, we develop high performance algorithms for certain computations involving dense tensors and sparse matrices. We address kernel operations that are useful for machine learning tasks, such as inference with deep neural networks (DNNs). We develop data structures and techniques to reduce memory use, to improve data locality and hence to improve cache reuse of the kernel operations. We design both sequential and shared-memory parallel algorithms. In the first part of the thesis we focus on dense tensors kernels. Tensor kernels include the tensor--vector multiplication (TVM), tensor--matrix multiplication (TMM), and tensor--tensor multiplication (TTM). Among these, TVM is the most bandwidth-bound and constitutes a building block for many algorithms. We focus on this operation and develop a data structure and sequential and parallel algorithms for it. We propose a novel data structure which stores the tensor as blocks, which are ordered using the space-filling curve known as the Morton curve (or Z-curve). The key idea consists of dividing the tensor into blocks small enough to fit cache, and storing them according to the Morton order, while keeping a simple, multi-dimensional order on the individual elements within them. Thus, high performance BLAS routines can be used as microkernels for each block. We evaluate our techniques on a set of experiments. The results not only demonstrate superior performance of the proposed approach over the state-of-the-art variants by up to 18%, but also show that the proposed approach induces 71% less sample standard deviation for the TVM across the d possible modes. Finally, we show that our data structure naturally expands to other tensor kernels by demonstrating that it yields up to 38% higher performance for the higher-order power method. Finally, we investigate shared-memory parallel TVM algorithms which use the proposed data structure. Several alternative parallel algorithms were characterized theoretically and implemented using OpenMP to compare them experimentally. Our results on up to 8 socket systems show near peak performance for the proposed algorithm for 2, 3, 4, and 5-dimensional tensors. In the second part of the thesis, we explore the sparse computations in neural networks focusing on the high-performance sparse deep inference problem. The sparse DNN inference is the task of using sparse DNN networks to classify a batch of data elements forming, in our case, a sparse feature matrix. The performance of sparse inference hinges on efficient parallelization of the sparse matrix--sparse matrix multiplication (SpGEMM) repeated for each layer in the inference function. We first characterize efficient sequential SpGEMM algorithms for our use case. We then introduce the model-parallel inference, which uses a two-dimensional partitioning of the weight matrices obtained using the hypergraph partitioning software. The model-parallel variant uses barriers to synchronize at layers. Finally, we introduce tiling model-parallel and tiling hybrid algorithms, which increase cache reuse between the layers, and use a weak synchronization module to hide load imbalance and synchronization costs. We evaluate our techniques on the large network data from the IEEE HPEC 2019 Graph Challenge on shared-memory systems and report up to 2x times speed-up versus the baseline
Gilbert, Frédéric. "Méthodes et modèles pour la visualisation de grandes masses de données multidimensionnelles nominatives dynamiques." Thesis, Bordeaux 1, 2012. http://www.theses.fr/2012BOR14498/document.
Full textSince ten years, informations visualization domain knows a real interest.Recently, with the growing of communications, the research on social networks analysis becomes strongly active. In this thesis, we present results on dynamic social networks analysis. That means that we take into account the temporal aspect of data. We were particularly interested in communities extraction within networks and their evolutions through time. [...]
Argoubi, Majdi El. "Optimisation de la gestion d’un réseau hydrographique en périodes de crues." Thesis, Ecole centrale de Lille, 2009. http://www.theses.fr/2009ECLI0002.
Full textA decision support system (DSS) has been developed for integrated, real-time flood control in a multi-reservoir system. The DSS integrates a geographic information system (GIS) with an optimization module, and a graphical dialog interface allowing effective use by system operators. A neuro-dynamic programming technique (NDP), which overcomes numerous limitations of dynamic programming (DP), is introduced for determining the optimal release policy under flood conditions. Results obtained in a real-world case study are finally presented
Ruty, Guillaume. "Towards more scalability and flexibility for distributed storage systems." Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLT006/document.
Full textThe exponentially growing demand for storage puts a huge stress on traditionnal distributed storage systems. While storage devices' performance have caught up with network devices in the last decade, their capacity do not grow as fast as the rate of data growth, especially with the rise of cloud big data applications. Furthermore, the performance balance between storage, network and compute devices has shifted and the assumptions that are the foundation for most distributed storage systems are not true anymore. This dissertation explains how several aspects of such storage systems can be modified and rethought to make a more efficient use of the resource at their disposal. It presents an original architecture that uses a distributed layer of metadata to provide flexible and scalable object-level storage, then proposes a scheduling algorithm improving how a generic storage system handles concurrent requests. Finally, it describes how to improve legacy filesystem-level caching for erasure-code-based distributed storage systems, before presenting a few other contributions made in the context of short research projects
Alleg, Abdelhamid. "Service Function Placement and Chaining in Network Function Virtualization Environments." Thesis, Bordeaux, 2019. http://www.theses.fr/2019BORD0117.
Full textThe emergence of Network Function Virtualization (NFV) technology has aroused keen interest to design, manage and deploy network services in a flexible, automated and vendor-agnostic manner. Implementing NFV technology is expected to be a win-win solution for both service providers and costumers. However, this paradigm shift, sparked by NFV, calls for a progressive abandon of network services that are provided as hardware appliance and rather it proposes a fully or partially virtualized environment that offers software modules called Virtual Network Functions (VNFs). This shift rises a set of challenges related to service deployment and operation such as orchestration and management, service resiliency, Quality of Service (QoS) and resource provisioning among others. Furthermore, the core question that needs to be solved within NFV context is “What is the best way to place and chain VNFs that form a service in order to meet Service Level Agreement requirements (costumer side) while optimizing resource usage (service provider side)?”.This thesis investigates the problem of VNF Placement and Chaining considering service requirements such as end-to-end delay, service availability and energy consumption and proposes a set of algorithms and mechanisms that aim to achieve an optimized deployment of the requested/provided services. Our contributions in this thesis are threefold. First, we propose a delay-aware Placement and Chaining algorithms for delay-sensitive applications over NFV networks. The proposed algorithms aim to meet the appropriate end-to-end delay defined according to the deployed service (VoIP, Streaming, etc.). Second, we provide a comprehensive service availability benchmarking and we propose two availability-aware mechanisms for VNFs chain. The aim is to provide resilient service provisioning by fine-tuning the parameters of the protection scheme (the number, the type, the placement and the size of the spare instances) needed to reach a predefined availability level, despite network failures. Finally, we propose a framework architecture that explores the possibility to extend the virtualization paradigm to Internet of Things (IoT). Toward this end, we define an energy-aware Placement and Chaining for IoT services where inherent IoT functionalities are decoupled from specific dedicated IoT devices and instantiated on-demand. By bringing together NFV and IoT paradigms, this extension opens new perspectives and push toward designing new use cases
Ghorbel, Asma. "Limites Fondamentales De Stockage Dans Les Réseaux Sans Fil." Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLC031/document.
Full textCaching, i.e. storing popular contents at caches available at end users, has received a significant interest as a technique to reduce the peak traffic in wireless networks. In particular, coded caching proposed by Maddah-Ali and Niesen has been considered as a promising approach to achieve a constant delivery time as the dimension grows. However, several limitations prevent its applications in practical wireless systems. Throughout the thesis, we address the limitations of classical coded caching in various wireless channels. Then, we propose novel delivery schemes that exploit opportunistically the underlying wireless channels while preserving partly the promising gain of coded caching. In the first part of the thesis, we study the achievable rate region of the erasure broadcast channel with cache and state feedback. We propose an achievable schemeand prove its optimality for special cases of interest. These results are generalized to the multi-antenna broadcast channel with state feedback. In the second part, we study the content delivery over asymmetric block-fading broadcast channels, where the channel quality varies across users and time. Assuming that user requests arrive dynamically, we design an online scheme based on queuing structure and prove that it maximizes the alpha-fair utility among all schemes restricted to decentralized placement. In the last part, we study opportunistic scheduling over the asymmetric fading broadcast channel and aim to design a scalable delivery scheme while ensuring fairness among users. We propose a simple threshold-based scheduling policy of linear complexity that requires only a one-bit feedback from each user
Lardas, Menelaos. "Modélisation de l'atelier flexible avec capacité limitée de stockage." Paris 11, 1986. http://www.theses.fr/1986PA112134.
Full textLn this thesis we consider the performance evaluation of flexible manufacturing systems. We study a flexible manufacturing system with three servers within a station. We model the behavior of this system with limited or unlimited capacity queues. Based on queueing theory an analytic model is found and is validated by simulations. Finally, we use the same approach to cope with the problem of a flexible manufacturing system with N stations in tandem with limited or unlimited capacity
Devigne, Julien. "Protocoles de re-chiffrement pour le stockage de données." Caen, 2013. http://www.theses.fr/2013CAEN2032.
Full textPrivacy is one of the main issues of our modern day society in which the Internet is omnipotent. In this thesis, we study some technics allowing to realise a privacy-preserving cloud storage. In this way, we focus to protect stored data while allowing their owner to share them with people of his choice. Proxy re-encryption, one of the primitives offered by cryptography, is the solution we decide to consider. First, we give a definition of a proxy re-encryption system unifying all existing conventional models. We also describe usual characteristics that this primitive may present and we provide its security model. Then, we focus more precisely on some specific schemes in order to improve their security. In this meaning, we expose a method which turns a scheme secure against a replayable chosen ciphertext attack into a secure scheme against a chosen ciphertext attack. We study schemes based on the Hash ElGamal encryption too and propose some modifications in order to reach a better security. Finally and in order to obtain the most functional cloud storage, we propose two new models. The first one, that we call combined proxy re-encryption, offers dynamic right access. The second one, that we call selective proxy re-encryption, enables a more fine-grained access right control than the one offered by the conditional proxy re-encryption
Bouabache, Fatiha. "Stockage fiable des données dans les grilles, application au stockage des images de checkpoint." Paris 11, 2010. http://www.theses.fr/2010PA112329.
Full textRollback/recovery solutions rely on checkpoint storage reliability (after a failure, if the checkpoint images are not available, the rollback operation fails). The goal of this thesis is to propose a reliable and an efficient checkpoint storage service. By reliable, we mean that whatever the scenario of failures is, as long as it respects the assumptions made by the algorithms, the checkpoint images are still available. And we mean by efficient, minimizing the time required to transfer and to store the checkpoint images. This will minimize the global execution time of the checkpoint waves. To ensure those two points (reliability and efficiency), we propose: 1. A new coordinated checkpoint protocol which tolerates checkpoint server failures and clusters failures, and ensures a checkpoint storage reliability in a grid environment; 2. A distributed storage service structured on three layers architecture: a) The replication layer: to ensure the checkpoint storage reliability, we propose to replicate the images over the network. Ln this direction, we propose two hierarchical replication strategies adapted to the considered architecture and that exploit the locality of checkpoint images in order to minimize inter-cluster communication. B) The scheduling layer: at this level we work on the storage efficiency by reducing the data transfer time. We propose an algorithm based on the uniform random sampling of possible schedules. C) The scheduling engine: at this layer, we develop a tool that implements the scheduling plan calculated in the scheduling layer
Mongkoltanatas, Jiravan. "Participation d'un système de stockage à la stabilité des réseaux insulaires." Thesis, Grenoble, 2014. http://www.theses.fr/2014GRENT115/document.
Full textFrequency of isolated microgrids is highly sensitive to active power variation of loads and productions because of the inertia equivalent of this grid is small (limited number of thermal generators). Furthermore, the increasing of renewable energy in this grid causes frequency more risky to be instable because of its intermittency of power. In this thesis, energy storage is the selected solution to maintain the frequency stability of isolated microgrids with high penetration rate of renewable energy. It will participate to the primary frequency control which is the first control that takes action to limit the frequency deviation after disturbance. Therefore, the objective of this thesis is to design the appropriate size and strategy of energy storage in isolated microgrids with high penetration rate of photovoltaic. Proposed strategies will link to power variation of PV. Therefore, the different situations of photovoltaic power variations and their impacts on the system frequency had been firstly characterized and studied in order to specify the critical situation of frequency variation. Then, different strategies for energy storage and their sizes were defined from these two studies. The strategy limit dPpv diagram and filter strategy were proposed to define the participated power of energy storage in primary frequency control from the measured power variation of PV. Furthermore, controller H infinity which is robust control was also proposed. Finally, energy storage control system was validated by RTLAB (real time simulation) which enable us to simulate the hardware. The results show that energy storage by proposed strategies is able to stabilize frequency of the power system by limiting the frequency deviation to be within an acceptable range after occurrence of any disturbance. The proposed strategies can increase larger participated power of the energy storage with less frequently than the classic droop control
Martinelli, Matthieu. "Stockage d’énergie thermique par changement de phase – Application aux réseaux de chaleur." Thesis, Université Grenoble Alpes (ComUE), 2016. http://www.theses.fr/2016GREAI084/document.
Full textThis study is about a shell and tubes latent heat thermal energy storage system. This system is expected to be integrated in a district heating network substation. Heat transfers inside the PCM as well as convection flow regime inside the heat transfer fluid are investigated.A first experimental study aims at demonstrating the necessity of internal insert inside the tubes in order to avoid mixed convection flow regime. Two highly finned tubes as well as two inserts are tested. Inserts are either cylindrical or helical. Better thermal performances are obtained with the helical one. Besides, it is shown that free convection, between the fins is negligible. Effective thermal conductivities are estimated with an experimental and analytical approach at 7.4 and 10.9 W/m/K for the 7 fpi and the 10 fpi tube.A second test campaign is carried out with metallic foams. The first one is stochastic and in copper while the second one is regular and in alumina. Effective thermal conductivities are around 13.4 and 39.4 W/m/K respectively. The copper foam heat exchanger is shown to be better than a copper finned tube in terms of stored energy and thermal power, whereas only half the mass of the fins is used in the foam.Eventually a CFD numerical model is experimentally validated. This model shows that free convection inside the PCM is negligible on the overall thermal performances even though it modifies the solid/liquid interface shape locally
Sené, Sylvain. "Sur la bio-informatique des réseaux d'automates." Habilitation à diriger des recherches, Université d'Evry-Val d'Essonne, 2012. http://tel.archives-ouvertes.fr/tel-00759287.
Full textRibeiro, Cardoso André. "Architecture basée sur les réseaux programmables et les réseaux pair-à-pair." Paris, ENST, 2007. http://www.theses.fr/2007ENST0038.
Full textRossignol, Alexandre. "Stockage d'électricité et système électrique en mutation." Electronic Thesis or Diss., Université Paris sciences et lettres, 2023. http://www.theses.fr/2023UPSLD043.
Full textIn addition to facilitating the development of renewable energies, electricity storage helps to strengthen the security of electricity supply. By shifting electricity consumption over time, storage makes it possible to better correlate demand with supply. It also contributes to the diversification of flexibilities, necessary due to the intermittency, variability and decentralized nature of renewable electricity production. Until 2019, there was no legal framework governing the electricity storage activity. For convenience, the public authorities qualified the storage operator as both a producer and an end consumer of electricity. But the energy policy choices, mainly focused on the development of intermittent renewable electricity production, gave renewed interest to electricity storage from the mid-2010s, which led the European legislator to create a legal framework specific to the storage activity within the framework of the fourth energy “package”. Although the new European legislative framework for the internal electricity market has introduced several provisions referring to electricity storage, the current French legal framework remains insufficient, with certain European provisions also being insufficient
Fuin, David. "Qualité de service : des réseaux IP à l'intégration dans les réseaux actifs." Besançon, 2004. http://www.theses.fr/2004BESA2053.
Full textHohwiller, Luc. "Méthodologie pour la simulation de réseaux de terrain : application aux réseaux PROFIBUS-DP." Mulhouse, 2001. http://www.theses.fr/2001MULH0683.
Full textAbou, Chacra Fouad. "Valorisation et optimisation du stockage d'énergie dans un réseau d'énergie électrique." Paris 11, 2005. http://www.theses.fr/2005PA112063.
Full textFor more than a century, electric transmission and distribution systems have been developed assuming that electric energy was almost impossible to store. Technical progress, new environmental requirements and electrical industry reforms now lead us to believe that storage in the future will be one of the main challenges in the development of power systems. Storage would have potential applications to deal with current technical constraints such as the system load, peakload value, faults in parts of the system, control issues, etc. And economic ones such as upgrades deferral, renewable energy deployment, etc. In this study, energy storage is considered in two startegic locations in the French power system: HT/MT substations and wind farms. Possible applications and economic flags are formulated and appropriate optimization methods (genetic algorithms, Pareto) are used to maximize the project net present value. This optimization results in defining optimal capacities and control strategies for the energy storage system, taken from a set of storage technologies suitable for this problem, and in assessing the technico-economic impact of energy storage as a solution in power systems
Hadj, said Ahmed. "Intégration du stockage dans les méthodes de planification des réseaux électriques basse tension." Thesis, Université Grenoble Alpes (ComUE), 2018. http://www.theses.fr/2018GREAT004/document.
Full textThe opening up of energy markets and new uses have led to significant changes in distribution grids, in particular low-voltage grids. Notably, it has led to an augmentation in the integration of renewable energy production, an increase in the peak consumption, among others. This is accompanied by the appearance of the electrical constraints with which power systems must cope. This has resulted in the development multiple flexibility capabilities such as load/source management or energy storage, providing new solutions, now to be considered in planning methods. This thesis studies the issue of energy storage in the low-voltage grid planning. The first part of this thesis studies the impact of storage and photovoltaic production on variables involved in distribution grid planning. In the second part, a method for calculating the cost of losses is adapted to the presence of energy storage and/or PV production. Finally, advanced d operation algorithms are developed to illustrate the economic value of energy storage in LV distribution grid planning, compared to a more expensive conventional planning method
Cormier, Alexandre. "Modélisaton et sécurité des réseaux." Thesis, Université Laval, 2007. http://www.theses.ulaval.ca/2007/25012/25012.pdf.
Full textIrisson, Olivier. "L'entreprise en réseau : métaphore informatique et utopie organisationnelle." Paris 9, 2000. https://portail.bu.dauphine.fr/fileviewer/index.php?doc=2000PA090019.
Full textPertin, Dimitri. "Code à effacement Mojette pour le stockage distribué." Nantes, 2016. https://archive.bu.univ-nantes.fr/pollux/show/show?id=dec41aca-365a-4ffe-8bf0-578d2f878c68.
Full textErasure codes can generate data redundancy in distributed storage systems. This redundancy can be used to recover missing data in case of a failure. Codes have the benefit of reducing the generated amount of redundancy drastically, compared to plain data replication. However, this reduction is combined with a significant computational complexity, which penalizes encoding and decoding performances, and limits the use of coding to cold data. In this thesis, we focus on the use of the Mojette transform as an effective erasure code, adapted to hot data. The resulting code requires more redundancy than classical codes though. The first contribution of this research work deals with the design of a systematic version of the Mojette erasure code. This version provides better performances while reducing the required amount of redundancy. The second contribution covers the integration of this solution in the distributed file system RozoFS. This integration enables the system to provide a continuous service despite failures, while being able to manage hot data with half the volume of data compared to replication-based systems. A third research focus addresses the design of a distributed method to compute extra Mojette codeword symbols. This method contributes to restore a redundancy threshold in the storage system
Giannakas, Theodoros. "Joint modeling and optimization of caching and recommendation systems." Electronic Thesis or Diss., Sorbonne université, 2020. http://www.theses.fr/2020SORUS317.
Full textCaching content closer to the users has been proposed as a win-win scenario in order to offer better rates to the users while saving costs from the operators. Nonetheless, caching can be successful if the cached files manage to attract a lot of requests. To this end, we take advantage of the fact that the internet is becoming more entertainment oriented and propose to bind recommendation systems and caching in order to increase the hit rate. We model a user who requests multiple contents from a network which is equipped with a cache. We propose a modeling framework for such a user which is based on Markov chains and depart from the IRM. We delve into different versions of the problem and derive optimal and suboptimal solutions according to the case we examine. Finally we examine the variation of the Recommendation aware caching problem and propose practical algorithms that come with performance guarantees. For the former, the results indicate that there are high gains for the operators and that myopic schemes without a vision, are heavily suboptimal. While for the latter, we conclude that the caching decisions can significantly improve when taking into consideration the underlying recommendations
Carrié, Stéphanie. "Le livre à l'épreuve des réseaux." Montpellier 1, 2005. http://www.theses.fr/2005MON10022.
Full textVoegtlin, Thomas. "Réseaux de neurones et auto-référence." Lyon 2, 2002. http://theses.univ-lyon2.fr/documents/lyon2/2002/voegtlin_t.
Full textThe purpose of this thesis is to present a class of unsupervised learning algorithms for recurrent networks. In the first part (chapters 1 to 4), I propose a new approach to this question, based on a simple principle: self-reference. A self-referent algorithm is not based on the minimization of an objective criterion, such as an error function, but on a subjective function, that depends on what the network has previously learned. An example of a supervised recurrent network where learning is self-referent is the Simple Recurrent Network (SRN) by Elman (1990). In the SRN, self-reference is applied to the supervised error back-propagation algorithm. In this aspect, the SRN differs from other generalizations of back-propagation to recurrent networks, that use an objective criterion, such as Back-Propagation Through Time, or Real-Time Recurrent Learning. In this thesis, I show that self-reference can be combined with several well-known unsupervised learning methods: the Self-Organizing Map (SOM), Principal Components Analysis (PCA), and Independent Components Analysis (ICA). These techniques are classically used to represent static data. Self-reference allows one to generalize these techniques to time series, and to define unsupervised learning algorithms for recurrent networks
Le, Hung-Cuong. "Optimisation d'accès au médium et stockage de données distribuées dans les réseaux de capteurs." Besançon, 2008. http://www.theses.fr/2008BESA2052.
Full textWireless sensor network is a very hot research topic tendency for the last few years. This technology can be applied into different domains as environment, industry, commerce, medicine, military etc. Depending on the application type, the problems and requirements might be different. In this thesis, we are interested in two major problems: the medium access control and the distributed data storage. The document is divided to two parts where the first part is a state of the art of different existing works and the second part describes our contribution. In the first contribution, we have proposed two MAC protocols. The first one optimizes the wireless sensor networks lifetime for surveillance applications and the second one reduces the transmission latency in event-driven wireless sensor networks for critical applications. In the second contribution, we have worked with several data storage models in wireless sensor network and we focus on the data-centric storage model. We have proposed a clustering structure for sensors to improve the routing and reduce the number of transmissions in order to prolong the network lifetime
Lagacherie, Hervé. "L'analyse des données cliniques et biologiques par les réseaux neuronaux." Bordeaux 2, 1997. http://www.theses.fr/1997BOR2P091.
Full textRonald, Edmund. "Apprentissage évolutionniste des réseaux neuromimétiques." Palaiseau, Ecole polytechnique, 1997. http://www.theses.fr/1997EPXX0048.
Full textPopa, Daniel. "Performance issues in metropolitan optical networks : packet format, MAC protocol and QoS." Evry, Institut national des télécommunications, 2005. http://www.theses.fr/2005TELE0002.
Full textThe key problems in networking in metropolitan area are flexibility and performance at MAC level. In this context, we identify the simpler and the faster solution to deliver the client layer information in presence of different QoS demands in metropolitan optical networks, employing ring topology and optical CSMA/CA protocol. The problem of choosing an optical packet format (i. E. Fixed or variable is extremely delicate because a large array of arguments for a response to the problem of electing the best approach based on the analysis of interaction between electrical to optical interface and packet format as well as the influence of optical packet format on access protocol performance. Since one of our objectives is to design an access protocol with QoS capabilities, the important aspects of fairness and QoS introduction in metropolitan optical networks are explored and new solutions are designed to deal with these problems. Firstly, we propose two mechanisms, named Glue MAC and SPACES (Smart spACEd tranSmission), Which efficiently prevent from bandwidth fragmentation and positional priority fairness problems. Next, we introduce a quantitative measure of fairness distribution among access points in ring topologies. Finally, we design a hybrid access scheme which combines features of CSMA/CA scheme, and a transparent transport of TDM-based flows
Nguyen, Hoai-Tuong. "Réseaux bayésiens et apprentissage ensembliste pour l'étude différentielle de réseaux de régulation génétique." Phd thesis, Université de Nantes, 2012. http://tel.archives-ouvertes.fr/tel-00675310.
Full textBigot, Pascal. "Utilisation des réseaux de neurones pour la télégestion des réseaux techniques urbains." Lyon 1, 1995. http://www.theses.fr/1995LYO10036.
Full text