Tesis sobre el tema "Adaptation de réseaux"
Crea una cita precisa en los estilos APA, MLA, Chicago, Harvard y otros
Consulte los 50 mejores tesis para su investigación sobre el tema "Adaptation de réseaux".
Junto a cada fuente en la lista de referencias hay un botón "Agregar a la bibliografía". Pulsa este botón, y generaremos automáticamente la referencia bibliográfica para la obra elegida en el estilo de cita que necesites: APA, MLA, Harvard, Vancouver, Chicago, etc.
También puede descargar el texto completo de la publicación académica en formato pdf y leer en línea su resumen siempre que esté disponible en los metadatos.
Explore tesis sobre una amplia variedad de disciplinas y organice su bibliografía correctamente.
Vantomme, Ghislaine. "Réseaux dynamiques constitutionnels : métallosélection - photosélection - adaptation". Phd thesis, Université de Strasbourg, 2014. http://tel.archives-ouvertes.fr/tel-01060251.
Texto completoYawut, Cholatip. "Adaptation à la mobilité dans les réseaux ad hoc". Thesis, Toulouse, INPT, 2009. http://www.theses.fr/2009INPT036H/document.
Texto completoManquant
Hua, Fei. "Adaptation et apprentissage sur des réseaux et des graphiques multitâches". Thesis, Université Côte d'Azur, 2020. http://www.theses.fr/2020COAZ4037.
Texto completoMultitask learning has received considerable attention in signal processing and machine learning communities. It aims at simultaneously learning several related tasks other than the traditional single-task problems. There also have witnessed a wide spectrum of data processing problems that are network- or graph-structured and require adaptation ability to streaming data and time-varying dynamics. Distributed adaptive learning strategies over networks enable a collection of interconnected agents to accomplish a certain task, such as parameter estimation, collaboratively through local computation and cooperation among neighboring agents. Further, they endow the agents with continuous adaptation and learning ability to track possible drifts in the underlying model. Despite the heterogeneous nature and the fact that each agent may solve a different task in multitask network, it could still benefit from a collaboration between agents to improve the estimation accuracy by leveraging the relations and capitalizing on inductive transfer between them. The objective of this thesis is to devise and analyze multitask adaptive learning strategies over networks and graphs. First, we consider multitask estimation problems where each agent is interested in estimating its own parameter vector and where the parameter vectors at neighboring agents are related linearly according to a set of constraints. Based on the penalty method, an unconstrained optimization problem is reformulated and a distributed algorithm is derived. The behavior of the algorithm in the mean and in the mean-square-error sense is analyzed. Next, we relax the local constraints assumption and consider the multitask problem with non-local constraints. We devise the distributed algorithm by employing a multi-hop relay protocol across the agents. We prove that the algorithm will continue to converge and provide theoretical performance analysis. In the third part, we extend the distributed learning strategies to the emerging graph signal processing applications where the signal itself is network-structured. Several graph diffusion LMS strategies are proposed to cope with streaming graph signals. We also extend the multitask model to graph filters and propose an on-line clustering mechanism. Last, we consider the problem of modeling graph signals by using a combination of multiple graph filters. An efficient algorithm is proposed to simultaneously learn coefficients of multiple graph filters and perform model selection. Simulation and numerical results are provided to illustrate the effectiveness of all proposed algorithms and validate the theoretical analyses
Abdallah, Alaeddine. "Mécanismes Cross-Layer pour le streaming vidéo dans les réseaux WIMAX". Thesis, Bordeaux 1, 2010. http://www.theses.fr/2010BOR14142/document.
Texto completoDriven by the increasing demand for multimedia services in broadband Internet networks, WIMAX technology has emerged as a competitive alternative to the wired broadband access solutions. The IEEE 802.16 is a solution that provides high throughput by ensuring a satisfactory QoS. In particular, it is suitable for multimedia applications that have strict QoS constraints. However, the users’ heterogeneity and diversity in terms of bandwidth, radio conditions and available resources, pose new deployment challenges. Indeed, multimedia applications need to interact with their environment to inform the access network about their QoS requirements and dynamically adapt to changing network conditions.In this context, we propose two solutions for video streaming over 802.16 networks based on Cross-Layer approach. We are interested in both unicast and multicast transmissions in uplink and downlink of one or more WIMAX cells.First, we proposed an architecture that enables Cross-Layer adaptation and optimization of video streaming based on available resources. We defined the entity CLO (Cross-Layer Optimizer) that takes benefits from service flow management messages, exchanged between BS and SS, at the MAC level, to determine the necessary adaptations / adjustment to ensure optimal delivery of the application. Adaptations occur at two epochs, during the admission of the video stream and during the streaming phase. The performance analysis, performed through simulations, shows the effectiveness of the CLO to adapt in a dynamic way, the video data rate depending on network conditions, and thus guarantee an optimal QoS.Second, we proposed a solution that enables IP multicast video delivery in WIMAX network. This solution allows finding the compromise between the diversity of end-user requirements, in terms of radio conditions, modulation schemes and available resources, along with the SVC hierarchy video format, to offer the best video quality even for users with low radio conditions. Indeed, we define a multicast architecture that allows each user to get a video quality proportionally to its radio conditions and its available bandwidth. Towards this end, several IP multicast groups are created depending on the SVC video layers. Subsequently, our solution allows optimizing the use of radio resources by exploiting the different modulations that can be selected by the end-users
Daucé, Emmanuel. "Adaptation dynamique et apprentissage dans les réseaux de neurones récurrents aléatoires". Toulouse, ENSAE, 2000. https://tel.archives-ouvertes.fr/tel-01394004.
Texto completoMezghani, Manel. "Analyse des réseaux sociaux : vers une adaptation de la navigation sociale". Thesis, Toulouse 3, 2015. http://www.theses.fr/2015TOU30127/document.
Texto completoThe advent of Web 2.0, user-centered, has given rise to a significant amount of information (personal, collective, shared, "loved", etc.). This information is a way to help users and guide them to the information sought. However, this quantity makes access to shared information more and more difficult, given the diversity of content that may interest the user. Disorientation of the user is one of the main problems related to social media. To overcome such problem, adaptation is a standard solution that can be applied in a social context. With the evolution of these social networks, new concepts appear such as social navigation, which is a way to navigate while being influenced by other users in the network: Another important concept is that of "tag". This term is defined as social annotations created by users and associated to resources. Navigation can be therefore carried out by both links and tags. Adapting social navigation means making it more targeted for each user according to their interests. In practice, this can be done by recommending tags to each user, so he can follow or not. To adapt the social navigation, we must ensure proper detection of the user's interests and taking into account their evolution. However, we are faced with some problems: i) the detection of interest, since they can be derived from several social resources (friends, resources, tags, etc.). Their relevance is primordial to ensure adequate adaptation result. ii) updating the user profile. Indeed, the social user, is characterized by its great social activity, and therefore its interests should reflect its "real" interest each time period in order to achieve a reliable adaptation. To solve the problems affecting the quality of adaptation of social navigation quoted above, we first proposed a method for detecting the user's interests. This proposal aims to overcome the detection of irrelevant interests issues. This approach analyzes the user tags depending on the content of their respective resources. Unlike most research, who do not consider the accuracy of tags with the contents of resource, the accuracy reflects whether the user is really interested with the content or not. This is done by querying the user's network and analysis of the user annotation behavior. The approach is based on the assumption that a user annotates the resource by tags reflecting the content of this resource better reflects its "true" interests. Following the proposal of the interests of detection approach, we conducted second, the treatment of the problem of updating these interests. We were interested to the user profile enrichment techniques, performed by adding interests deemed relevant at a given time. The enrichment in a social context is performed according to social information such as neighbours who share the user behaviors in common, according to the user annotation behavior, and according to the metadata annotated resources. The choice of such information shall follow the study of their influence on the changing interests of the user. The approach we used enrichment propose recommendations (tags) according to the new tags added to the user profile. Both contributions were tested on the social database Delicious. They showed a sizeable accuracy rate. They have also proven their efficiency compared to conventional methods. In addition, the rate of ambiguity associated with the tags has been greatly reduced, thanks to the implicit filtering of irrelevant tags relative to resource content
Chen, Jun. "Adaptation de liens et mécanisme inter-couches pour les réseaux coopératifs". Paris, CNAM, 2010. http://www.theses.fr/2010CNAM0702.
Texto completoLes techniques de transmission coopérative dans les réseaux sans fil concernent les méthodes dont l'objectif principal est d'améliorer la fiabilité de la transmission d'information entre la source et la destination en exploitant des terminaux intermédiaires appelés relais. L'objectif principal de cette thèse est de proposer un mécanisme efficace d'adaptation de liens en utilisant les techniques inter-couches dans les réseaux coopératifs. L'évaluation des performances des modules coopératifs, en utilisant soit un seul relais soit plusieurs relais, est présentée. Par ailleurs, la fiabilité de la transmission peut être améliore��e par le déploiement de codage espace-temps distribué (DSTC) dans la coopération. De plus, un mécanisme de sélection de relais et d'adaptation de mode physique a été développé pour améliorer le débit efficace pendant la transmission. En analysant la valeur du rapport signal à bruit (SNR) de bout en bout pour les transmissions coopératives, le protocole coopératif qui assure le taux optimal de transmission peut alors être déterminé. Finalement, grâce à une spécification de la couche MAC intégrant la gestion de la coopération, le choix du meilleur relais et l’adaptation de liens sont présentés. Cette thèse montre, d'une part, que la capacité de lien dans les réseaux coopératifs peut être améliorée lorsque les modes physiques du relais sont bien adaptés aux conditions des canaux, et d'autre part, que des stratégies inter-couches efficaces peuvent fournir un meilleur support de la qualité de service
Shehadeh, Dareen. "Dynamic network adaptation for energy saving". Thesis, Ecole nationale supérieure Mines-Télécom Atlantique Bretagne Pays de la Loire, 2017. http://www.theses.fr/2017IMTA0067/document.
Texto completoThe main goal of the thesis is to design an Energy Proportional Network by taking intelligent decisions into the network such as switching on and off network components in order to adapt the energy consumption to the user needs. Our work mainly focuses on reducing the energy consumption by adapting the number of APs that are operating to the actual user need. In fact, traffic load varies a lot during the day. Traffic is high in urban areas and low in the suburb during day work hours, while it is the opposite at night. Often, peak loads during rush hours are lower than capacities of the networks. Thus they remain lightly utilized for long periods of time. Thus keeping all APs active all the time even when the traffic is low causes a huge waste of energy. Our goal is to benefit from low traffic periods by automatically switch off redundant cells, taking into consideration the actual number of users, their traffic and the bandwidth requested to serve them. Ideally we wish to do so while maintaining reliable service coverage for existing and new coming users. First we consider a home networking scenario. In this case only one AP covers a given area. So when this AP is switched off (when no users are present), there will be no other AP to fill the gap of coverage. Moreover, upon the arrival of new users, no controller or other mechanism exists to wake up the AP. Consequently, new arriving users would not be served and would remain out of coverage. The study of the state of the art allowed us to have a clear overview of the existing approaches in this context. As a result, we designed a platform to investigate different methods to wake up an AP using different technologies. We measure two metrics to evaluate the Switching ON/OFF process for the different methods. The first is the energy consumed by the AP during the three phases it goes through. The second is the delay of time for the AP to wake up and be operational to serve the new users. In the second case we consider a dense network such as the ones found in urban cities, where the coverage area of an AP is also covered by several other APs. In other words, the gap resulting from switching off one or several APs can be covered by other neighbouring ones. Thus the first thing to do was to evaluate the potential of switching off APs using real measurements taken in a dense urban area. Based on this collected information, we evaluate how many APs can be switched off while maintaining the same coverage. To this end, we propose two algorithms that select the minimum set of APs needed to provide full coverage. We compute several performance parameters, and evaluate the proposed algorithms in terms of the number of selected APs, and the coverage they provide
Briche, Julien. "Adaptation d'un algorithme génétique pour la reconstruction de réseaux de régulation génétique : COGARE". Phd thesis, Université du Sud Toulon Var, 2009. http://tel.archives-ouvertes.fr/tel-00479671.
Texto completoBillet, Yves-Gaël. "Adaptation dynamique d'applications multimédia à leur contexte d'exécution dans les réseaux du futur". Phd thesis, Université Jean Monnet - Saint-Etienne, 2012. http://tel.archives-ouvertes.fr/tel-00955366.
Texto completoHua, Thanh Phong. "Adaptation mode controllers for adaptive microphone arrays". Rennes 1, 2006. http://www.theses.fr/2006REN1S136.
Texto completoSammarco, Matteo. "Dissémination multi-contenus opportuniste : monitorage passif et adaptation aux conditions du réseau". Thesis, Paris 6, 2014. http://www.theses.fr/2014PA066573/document.
Texto completoThe market penetration of mobile devices has experienced an impressive growth. Smartphones, tablets, and laptops have become both producers and consumers of user-generated contents. They also motivate novel communication paradigms such as the possibility to establish, in an opportunistic fashion, direct device-to-device links whenever two mobile nodes enter within the wireless range of each other. In this thesis, we consider the case of opportunistic dissemination of multiple large contents from an experimental point of view. This implies revisiting, among others, the common assumption that contacts have enough capacity to transfer any amount of data.In the first part of this thesis, we start from an Android implementation of EPICS, a network protocol designed for exchanging large contents in opportunistic networks, on off-the-shelf devices. After an deep analysis of application-level logs and captured wireless traces we found out limitations and uncovered improving possibilities. We then propose DAD, a new content dissemination protocol that adaptively sends bursts of data instead of the per-fragment transmission strategy of EPICS.The second part of this thesis deals with the scalability of legacy WLAN monitoring systems. We propose two original approaches. With the first one, based on trace similarity and community detection algorithms, we are able to identify how many monitor we need in a target area and where to place them. The second approach in based on collaborative measurements. In this case we face the risk of biased measures due attacks of malicious users generating adulterated traces. We then propose a method to detect such malicious behaviors
Cherif, Wael. "Adaptation de contexte basée sur la Qualité d'Expérience dans les réseaux Internet du Futur". Phd thesis, Université Rennes 1, 2013. http://tel.archives-ouvertes.fr/tel-00940287.
Texto completoSammarco, Matteo. "Dissémination multi-contenus opportuniste : monitorage passif et adaptation aux conditions du réseau". Electronic Thesis or Diss., Paris 6, 2014. http://www.theses.fr/2014PA066573.
Texto completoThe market penetration of mobile devices has experienced an impressive growth. Smartphones, tablets, and laptops have become both producers and consumers of user-generated contents. They also motivate novel communication paradigms such as the possibility to establish, in an opportunistic fashion, direct device-to-device links whenever two mobile nodes enter within the wireless range of each other. In this thesis, we consider the case of opportunistic dissemination of multiple large contents from an experimental point of view. This implies revisiting, among others, the common assumption that contacts have enough capacity to transfer any amount of data.In the first part of this thesis, we start from an Android implementation of EPICS, a network protocol designed for exchanging large contents in opportunistic networks, on off-the-shelf devices. After an deep analysis of application-level logs and captured wireless traces we found out limitations and uncovered improving possibilities. We then propose DAD, a new content dissemination protocol that adaptively sends bursts of data instead of the per-fragment transmission strategy of EPICS.The second part of this thesis deals with the scalability of legacy WLAN monitoring systems. We propose two original approaches. With the first one, based on trace similarity and community detection algorithms, we are able to identify how many monitor we need in a target area and where to place them. The second approach in based on collaborative measurements. In this case we face the risk of biased measures due attacks of malicious users generating adulterated traces. We then propose a method to detect such malicious behaviors
Nicolas, Charbel. "Sur l'adaptation au contexte des réseaux de capteurs sans fil". Thesis, Evry, Institut national des télécommunications, 2012. http://www.theses.fr/2012TELE0037/document.
Texto completoBeing mobile, the wireless sensors must adapt to the changing environment. Therefore, in the first part of this thesis we propose a mechanism to adapt the WSN architecture dynamically based on the detected context; this includes the dynamic detection of the topology change, the detection of the new context and consequently the dynamic adaptation of the communication layer. All of these actions are executed under constraints on energy consumption. The work done in this part poses the question of detecting the new context. This is a rather difficult question because it is unclear. The purpose of the second part is to detect on the fly the type of the competitor technology generating a traffic that interferes with the WSN. The proposed mechanism, FIM, identifies the cause of interference from errors model observed in the corrupt data packets. The context detection allows the nodes of the sensor network to obtain information about the environment. Some nodes must have more reliable information on the environment than others. How to retrieve the information? From which neighboring nodes? And what information to keep as safe and useful? Are the questions that are addressed in the third part. We propose a mechanism to dynamically decide if docition mechanisms should be used or not
Bestak, Robert. "Les mécanismes de fiabilisation (protocoles ARQ) et leur adaptation dans les réseaux radiomobiles de 3G". Phd thesis, Télécom ParisTech, 2003. http://tel.archives-ouvertes.fr/tel-00005738.
Texto completoHu, Ruijing. "Algorithmes de dissémination épidémiques dans les réseaux à grande échelle : comparaison et adaptation aux topologies". Phd thesis, Université Pierre et Marie Curie - Paris VI, 2013. http://tel.archives-ouvertes.fr/tel-00931796.
Texto completoSfar, Safouane. "Adaptation dynamique des services télécoms multimédias mobiles". Rennes 1, 2003. http://www.theses.fr/2003REN10067.
Texto completoTou, Ihsane. "Adaptation de la couche transport des systèmes de communication hybrides satellite/terrestre". Phd thesis, Université Paul Sabatier - Toulouse III, 2012. http://tel.archives-ouvertes.fr/tel-00807583.
Texto completoNicolas, Charbel. "Sur l'adaptation au contexte des réseaux de capteurs sans fil". Electronic Thesis or Diss., Evry, Institut national des télécommunications, 2012. http://www.theses.fr/2012TELE0037.
Texto completoBeing mobile, the wireless sensors must adapt to the changing environment. Therefore, in the first part of this thesis we propose a mechanism to adapt the WSN architecture dynamically based on the detected context; this includes the dynamic detection of the topology change, the detection of the new context and consequently the dynamic adaptation of the communication layer. All of these actions are executed under constraints on energy consumption. The work done in this part poses the question of detecting the new context. This is a rather difficult question because it is unclear. The purpose of the second part is to detect on the fly the type of the competitor technology generating a traffic that interferes with the WSN. The proposed mechanism, FIM, identifies the cause of interference from errors model observed in the corrupt data packets. The context detection allows the nodes of the sensor network to obtain information about the environment. Some nodes must have more reliable information on the environment than others. How to retrieve the information? From which neighboring nodes? And what information to keep as safe and useful? Are the questions that are addressed in the third part. We propose a mechanism to dynamically decide if docition mechanisms should be used or not
Hatoum, Rima. "Algorithmes d'ordonnancement inter-couches avec adaptation de modulation et de codage dans les réseaux hétérogènes LTE". Thesis, Paris 6, 2014. http://www.theses.fr/2014PA066159/document.
Texto completoResponding to the huge demand for high data rates and satisfying Quality of Service (QoS) requirements are the main objectives of the wireless mobile operators. LTE-based small cell system is a promising technology offering the required services, enhanced indoor coverage and increased system capacity. However, many challenges face the constructors and the operators for an effective deployment of these systems. In this thesis, in order to evolve such systems and adduce significant enhancement in terms of QoS and throughput, we adopt a complementary strategy based on both data link layer and physical layer.First, we propose, on the link layer level, two scheduling and joint resource allocation algorithm based on the Adaptive Modulation and Coding (AMC) and the power control mechanism for OFDMA-Downlink and SC-FDMA-Uplink connections respectively. Namely, they refer to: “Downlink AMC-QRAP” and “Uplink AMC-QRAP”. The joint adaptation of the transmission power and the Modulation and Coding Scheme reliably interacts with the link quality on each allocated sub-channel. Two distinct user categories are considered to differentiate between different QoS levels required. Accordingly, a linear optimization model is performed for the problem resolution. A clustering approach proposed in the literature has been used as a tradeoff between centralized and distributed schemes. Particularly, we used the spectrum sensing technique as proposed on the physical layer to detect surrounding transmissions.After that, we tackle the physical layer “signal processing” techniques as an LTE enhancement approach. Thus, two contributions based on the “wavelet transform” (WT) are proposed.The “wavelet-based OFDM” technique is firstly proposed as an alternative transmission mode for both downlink and uplink of the LTE and beyond wireless systems. In fact, theOFDM suffers from several limitations in terms of PAPR, spectral efficiency, synchronization cost and the inter-carrier interference. We proved that with substituting the Fourier transform by the wavelet transform in the OFDM technique, the different OFDM limitations can be significantly overcame. Second, we proposed an enhanced spectrum sensing approach based on WT tool that accurately delimits the occupied resource blocks in the whole spectrum. These physical layer enhancements help improving the scheduling algorithm for the uplink proposed in our second contribution.Extensive network simulations with different network densities have been conducted, using several metrics such as spectral efficiency, throughput satisfaction rate, user outage and transmission power. We have compared our methods to several existing works in the literature and proved the outperformance for the different considered metrics
Hatoum, Rima. "Algorithmes d'ordonnancement inter-couches avec adaptation de modulation et de codage dans les réseaux hétérogènes LTE". Electronic Thesis or Diss., Paris 6, 2014. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2014PA066159.pdf.
Texto completoResponding to the huge demand for high data rates and satisfying Quality of Service (QoS) requirements are the main objectives of the wireless mobile operators. LTE-based small cell system is a promising technology offering the required services, enhanced indoor coverage and increased system capacity. However, many challenges face the constructors and the operators for an effective deployment of these systems. In this thesis, in order to evolve such systems and adduce significant enhancement in terms of QoS and throughput, we adopt a complementary strategy based on both data link layer and physical layer.First, we propose, on the link layer level, two scheduling and joint resource allocation algorithm based on the Adaptive Modulation and Coding (AMC) and the power control mechanism for OFDMA-Downlink and SC-FDMA-Uplink connections respectively. Namely, they refer to: “Downlink AMC-QRAP” and “Uplink AMC-QRAP”. The joint adaptation of the transmission power and the Modulation and Coding Scheme reliably interacts with the link quality on each allocated sub-channel. Two distinct user categories are considered to differentiate between different QoS levels required. Accordingly, a linear optimization model is performed for the problem resolution. A clustering approach proposed in the literature has been used as a tradeoff between centralized and distributed schemes. Particularly, we used the spectrum sensing technique as proposed on the physical layer to detect surrounding transmissions.After that, we tackle the physical layer “signal processing” techniques as an LTE enhancement approach. Thus, two contributions based on the “wavelet transform” (WT) are proposed.The “wavelet-based OFDM” technique is firstly proposed as an alternative transmission mode for both downlink and uplink of the LTE and beyond wireless systems. In fact, theOFDM suffers from several limitations in terms of PAPR, spectral efficiency, synchronization cost and the inter-carrier interference. We proved that with substituting the Fourier transform by the wavelet transform in the OFDM technique, the different OFDM limitations can be significantly overcame. Second, we proposed an enhanced spectrum sensing approach based on WT tool that accurately delimits the occupied resource blocks in the whole spectrum. These physical layer enhancements help improving the scheduling algorithm for the uplink proposed in our second contribution.Extensive network simulations with different network densities have been conducted, using several metrics such as spectral efficiency, throughput satisfaction rate, user outage and transmission power. We have compared our methods to several existing works in the literature and proved the outperformance for the different considered metrics
Legtchenko, Sergey. "Adaptation dynamique des architectures réparties pour jeux massivement multijoueurs". Phd thesis, Université Pierre et Marie Curie - Paris VI, 2012. http://tel.archives-ouvertes.fr/tel-00931865.
Texto completoHabachi, Oussama. "Optimisation des Systèmes Partiellement Observables dans les Réseaux Sans-fil : Théorie des jeux, Auto-adaptation et Apprentissage". Phd thesis, Université d'Avignon, 2012. http://tel.archives-ouvertes.fr/tel-00799903.
Texto completoBeaudaux, Julien. "Auto-configuration et auto-adaptation de réseaux de capteurs sans fil dans le contexte de la télémédecine". Phd thesis, Université de Strasbourg, 2013. http://tel.archives-ouvertes.fr/tel-01015749.
Texto completoVu, Hien Duc. "Adaptation des méthodes d'apprentissage automatique pour la détection de défauts d'arc électriques". Electronic Thesis or Diss., Université de Lorraine, 2019. http://docnum.univ-lorraine.fr/ulprive/DDOC_T_2019_0152_VU.pdf.
Texto completoThe detection of electric arcs occurring in an electrical network by machine learning approaches represents the heart of the work presented in this thesis. The problem was first considered as a classification of fixed-size time series with two classes: normal and default. This first part is based on the work of the literature where the detection algorithms are organized mainly on a step of the transformation of the signals acquired on the network, followed by a step of extraction of descriptive characteristics and finally a step of decision. The multi-criteria approach adopted here aims to respond to systematic classification errors. A methodology for selecting the best combinations, transformation, and descriptors has been proposed by using learning solutions. As the development of relevant descriptors is always difficult, differents solutions offered by deep learning has also been studied. In a second phase, the study focused on the variable aspects in time of the fault detection. Two statistical decision paths have been explored, one based on the sequential probabilistic test (SPRT) and the other based on artificial neural networks LSTM (Long Short Time Memory Network). Each of these two methods exploits in its way the duration a first classification step between 0 and 1 (normal, default). The decision by SPRT uses an integration of the initial classification. LSTM learns to classify data with variable time. The results of the LSTM network are very promising, but there are a few things to explore. All of this work is based on experiments with the most complete and broadest possible data on the field of 230V alternative networks in a domestic and industrial context. The accuracy obtained is close to 100% in the majority of situations
NICOLAS, Charbel. "Sur l'adaptation au contexte des réseaux de capteurs sans fil". Phd thesis, Institut National des Télécommunications, 2012. http://tel.archives-ouvertes.fr/tel-00762223.
Texto completoNégru, Daniel. "Un Eco-système de Médias Réseaux basé sur la Conscience du Contexte, du Réseau et du Contenu à destination de l'Internet Media du Futur". Habilitation à diriger des recherches, Université Sciences et Technologies - Bordeaux I, 2013. http://tel.archives-ouvertes.fr/tel-01012046.
Texto completoMeftah, Sara. "Neural Transfer Learning for Domain Adaptation in Natural Language Processing". Thesis, université Paris-Saclay, 2021. http://www.theses.fr/2021UPASG021.
Texto completoRecent approaches based on end-to-end deep neural networks have revolutionised Natural Language Processing (NLP), achieving remarkable results in several tasks and languages. Nevertheless, these approaches are limited with their "gluttony" in terms of annotated data, since they rely on a supervised training paradigm, i.e. training from scratch on large amounts of annotated data. Therefore, there is a wide gap between NLP technologies capabilities for high-resource languages compared to the long tail of low-resourced languages. Moreover, NLP researchers have focused much of their effort on training NLP models on the news domain, due to the availability of training data. However, many research works have highlighted that models trained on news fail to work efficiently on out-of-domain data, due to their lack of robustness against domain shifts. This thesis presents a study of transfer learning approaches, through which we propose different methods to take benefit from the pre-learned knowledge on the high-resourced domain to enhance the performance of neural NLP models in low-resourced settings. Precisely, we apply our approaches to transfer from the news domain to the social media domain. Indeed, despite the importance of its valuable content for a variety of applications (e.g. public security, health monitoring, or trends highlight), this domain is still poor in terms of annotated data. We present different contributions. First, we propose two methods to transfer the knowledge encoded in the neural representations of a source model pretrained on large labelled datasets from the source domain to the target model, further adapted by a fine-tuning on few annotated examples from the target domain. The first transfers contextualised supervisedly pretrained representations, while the second method transfers pretrained weights, used to initialise the target model's parameters. Second, we perform a series of analysis to spot the limits of the above-mentioned proposed methods. We find that even if the proposed transfer learning approach enhances the performance on social media domain, a hidden negative transfer may mitigate the final gain brought by transfer learning. In addition, an interpretive analysis of the pretrained model, show that pretrained neurons may be biased by what they have learned from the source domain, thus struggle with learning uncommon target-specific patterns. Third, stemming from our analysis, we propose a new adaptation scheme which augments the target model with normalised, weighted and randomly initialised neurons that beget a better adaptation while maintaining the valuable source knowledge. Finally, we propose a model, that in addition to the pre-learned knowledge from the high-resource source-domain, takes advantage of various supervised NLP tasks
Michaut, Fabien. "Adaptation des applications distribuées à la Qualité de Service fournie par le réseau de communication". Nancy 1, 2003. http://docnum.univ-lorraine.fr/public/SCD_T_2003_0159_MICHAUT.pdf.
Texto completoVarious new networking applications have appeared in the past ten years. Packet switched networks initially designed for the transfer of text data are used nowadays by more complex traffics with strong constraints in terms of Quality of Service (QoS). Our approach considers that no strict guarantee can be obtained from communication system. In this context, we propose a new QoS Architecture called "QdS-Adapt". It offers a general framework that permits applications to adapt dynamically their execution to resources fluctuations. The architecture includes a network metrology service. It has been designed to allow the implementation of various measurement techniques. The feasibility of the architecture has been verified on two tele-operation applications of a mobile robot. Experimentations show that the architecture allows on-line adaptation strategies to be implemented and that the impact of QoS degradations on the application functioning can be limited
Ramadan, Wassim. "Amélioration de la transmission de contenus vidéo et de données dans les réseaux sans-fil". Phd thesis, Université de Franche-Comté, 2011. http://tel.archives-ouvertes.fr/tel-00802909.
Texto completoBoudour, Ghalem. "Bandwidth reservation in mobile ad hoc networks for providing QoS : adaptation for voice support". Toulouse 3, 2010. http://thesesups.ups-tlse.fr/1033/1/Boudour_Ghalem.pdf.
Texto completoQoS provisioning over Mobile Ad-Hoc Networks (MANETs) has attracted a great attention in recent years. While much research effort has been devoted to provide QoS over wired and cellular networks, QoS solutions for the support of real-time traffic over MANETs remains one of the most challenging and least explored areas. In fact, real-time applications such as voice and video could not function properly on MANETs without a QoS oriented medium access control (MAC) scheme. Indeed, real-time traffics claim strict requirements in terms of transmission delay and packet dropping that can be fulfilled only if the MAC sub-layer provides bounded channel access delay, and low collision rate. The purpose of this thesis is the proposal and analysis of an efficient reservation MAC protocol to provide QoS support over MANETs. Firstly, we study one major issue in resource reservation for MANETs which is reservation consistency. Our analysis of existing reservation MAC protocols for MANETs reveals that many reservation conflicts between neighbor nodes occur during the reservation establishment phase. These conflicts which are mainly due to collisions of reservation control messages, have an important impact on the performance of the reservation protocol, and lead to a significant collision and loss of packets during the life-time of the connection, which is not acceptable for real-time traffics. We design a new reservation MAC protocol that resolves these conflicts. The main principle of our protocol is to achieve better coordination between neighbor nodes in order to ensure consistency of reservations. Thus, before considering a reservation as successful, the protocol tries to ensure that each reservation control message transmitted by a node is successfully received by all its neighbors. In the second part of this thesis, we apply the proposed reservation protocol to voice traffic. .
Logiaco, Laureline. "Temporal modulation of the dynamics of neuronal networks with cognitive function : experimental evidence and theoretical analysis". Electronic Thesis or Diss., Paris 6, 2015. http://www.theses.fr/2015PA066225.
Texto completoWe investigated the putative function of the fine temporal dynamics of neuronal networks for implementing cognitive processes. First, we characterized the coding properties of spike trains recorded from the dorsal Anterior Cingulate Cortex (dACC) of monkeys. dACC is thought to trigger behavioral adaptation. We found evidence for (i) high spike count variability and (ii) temporal reliability (favored by temporal correlations) which respectively hindered and favored information transmission when monkeys were cued to switch the behavioral strategy. Also, we investigated the nature of the neuronal variability that was predictive of behavioral variability. High vs. low firing rates were not robustly associated with different behavioral responses, while deviations from a neuron-specific prototypical spike train predicted slower responses of the monkeys. These deviations could be due to increased or decreased spike count, as well as to jitters in spike times. Our results support the hypothesis of a complex spatiotemporal coding of behavioral adaptation by dACC, and suggest that dACC signals are unlikely to be decoded by a neural integrator. Second, we further investigated the impact of dACC temporal signals on the downstream decoder by developing mean-field equations to analyze network dynamics. We used an adapting single neuron model that mimics the response of cortical neurons to realistic dynamic synaptic-like currents. We approximated the time-dependent population rate for recurrent networks in an asynchronous irregular state. This constitutes an important step towards a theoretical study of the effect of temporal drives on networks which could mediate cognitive functions
Logiaco, Laureline. "Temporal modulation of the dynamics of neuronal networks with cognitive function : experimental evidence and theoretical analysis". Thesis, Paris 6, 2015. http://www.theses.fr/2015PA066225/document.
Texto completoWe investigated the putative function of the fine temporal dynamics of neuronal networks for implementing cognitive processes. First, we characterized the coding properties of spike trains recorded from the dorsal Anterior Cingulate Cortex (dACC) of monkeys. dACC is thought to trigger behavioral adaptation. We found evidence for (i) high spike count variability and (ii) temporal reliability (favored by temporal correlations) which respectively hindered and favored information transmission when monkeys were cued to switch the behavioral strategy. Also, we investigated the nature of the neuronal variability that was predictive of behavioral variability. High vs. low firing rates were not robustly associated with different behavioral responses, while deviations from a neuron-specific prototypical spike train predicted slower responses of the monkeys. These deviations could be due to increased or decreased spike count, as well as to jitters in spike times. Our results support the hypothesis of a complex spatiotemporal coding of behavioral adaptation by dACC, and suggest that dACC signals are unlikely to be decoded by a neural integrator. Second, we further investigated the impact of dACC temporal signals on the downstream decoder by developing mean-field equations to analyze network dynamics. We used an adapting single neuron model that mimics the response of cortical neurons to realistic dynamic synaptic-like currents. We approximated the time-dependent population rate for recurrent networks in an asynchronous irregular state. This constitutes an important step towards a theoretical study of the effect of temporal drives on networks which could mediate cognitive functions
Bouaziz, Hamida. "Adaptation of SysML Blocks and Verification of Temporal Properties". Thesis, Besançon, 2016. http://www.theses.fr/2016BESA2015/document.
Texto completoThe work presented in this thesis takes place in the component-based development domain, it is a contribution to the specification,adaptation and verification of component-based systems. The main purpose of this thesis is the proposition of a formal approach tobuild incrementally complex systems by assembling and adapting a set of components, where their structure and behaviour are modelledusing SysML diagrams. In the first stage, we have defined a meta-model driven approach which is based on meta-modelling and modelstransformation, to verify the compatibility of blocks having their interaction protocols modelled using SysML sequence diagrams. To verifytheir compatibility, we perform a transformation into interface automata (IAs), and we base on the optimistic approach defined on IAs. Thisapproach consider that two components are compatible if there is a suitable environment with which they can interact correctly. Afterthat, we have proposed to benefit from the hierarchy, that may be present in the interaction protocol models of the blocks, to alleviate theverification of blocks compatibility. In the next stage, we have taken into consideration the problem of names mismatches of type one2onebetween services of blocks. At this stage, an adapter is generated for a set of reused blocks which have their interaction protocols modelledformally by interface automata. The generation of the adapter is guided by the specification of the parent block which is made initiallyby the designer. Our approach is completed by a verification phase which allows us to verify SysML requirements, expressed formallyby temporal properties, on SySML blocks. In this phase, we have exploited only the generated adapters to verify the preservation of therequirements initially satisfied by the reused blocks. Thus, our approach intends to give more chance to avoid the state space explosionproblem during the verification. In the same context, where we have a set of reused blocks and the specification of their parent blocks, wehave proposed to use coloured Petri nets (CPNs) to model the blocks interactions and to generate adapters that solve more type of problems.In this case the adapter can solve the problem of livelock by enabling the reordering of services calls
El, Akchioui Nabil. "Fluidification des réseaux de Petri stochastiques : application aux études de fiabilité des systèmes". Le Havre, 2012. http://www.theses.fr/2012LEHA0014.
Texto completoReliability analysis based on discrete event systems and particularly on stochastic Petri nets, improves the safety of industrial processes and systems. For large scale systems, the fluidification of stochastic processes is useful to reduce the computational resources and also the duration of simulations. But global and direct fluidification that preserves the structure and the parameters of the original model leads to a biased estimation of the reliability indicators. This study explores innovative approaches to fluidify stochastic Petri nets. When the marking space has several regions, piecewise constant continuous Petri nets are introduced to divide the marking trajectory into several phases in order to reach a steady state identical with the steady state of the stochastic process. Homothetic and projective transformations are also proposed to reach a steady state that is partially homothetic to the one of the stochastic process. Finally, adaptive modifications of the parameters are used to correct progressively the mean marking. All proposed results contribute to design fluid models with asymptotic behaviours equivalent to the ones of a stochastic process
Lucidarme, Philippe. "Apprentissage et adaptation pour des ensembles de robots réactifs coopérants". Phd thesis, Université Montpellier II - Sciences et Techniques du Languedoc, 2003. http://tel.archives-ouvertes.fr/tel-00641563.
Texto completoKaddour, Mejdi. "Conception et réalisation d'une plate-forme de répartition dédiée aux environnements nomades". Phd thesis, Télécom ParisTech, 2005. http://pastel.archives-ouvertes.fr/pastel-00001225.
Texto completoMorvan, Ludivine. "Prédiction de la progression du myélome multiple par imagerie TEP : Adaptation des forêts de survie aléatoires et de réseaux de neurones convolutionnels". Thesis, Ecole centrale de Nantes, 2021. http://www.theses.fr/2021ECDN0045.
Texto completoThe aim of this work is to provide a model for survival prediction and biomarker identification in the context of multiple myeloma (MM) using PET (Positron Emission Tomography) imaging and clinical data. This PhD is divided into two parts: The first part provides a model based on Random Survival Forests (RSF). The second part is based on the adaptation of deep learning to survival and to our data. The main contributions are the following: 1) Production of a model based on RSF and PET images allowing the prediction of a risk group for multiple myeloma patients. 2) Determination of biomarkers using this model.3) Demonstration of the interest of PET radiomics.4) Extension of the state of the art of methods for the adaptation of deep learning to a small database and small images. 5) Study of the cost functions used in survival. In addition, we are, to our knowledge, the first to investigate the use of RSFs in the context of MM and PET images, to use self-supervised pre-training with PET images, and, with a survival task, to fit the triplet cost function to survival and to fit a convolutional neural network to MM survival from PET lesions
Mbaye, Maïssa. "Les systèmes cognitifs dans les réseaux autonomes : une méthode d'apprentissage distribué et collaboratif situé dans le plan de connaissance pour l'auto-adaptation". Thesis, Bordeaux 1, 2009. http://www.theses.fr/2009BOR13976/document.
Texto completoOne of the major challenges for decades to come, in the field of information technologies and the communication, is realization of autonomic paradigm. It aims to enable network equipments to self-manage, enable them to self-configure, self-optimize, self-protect and self-heal according to high-level objectives of their designers. Major architectures of autonomic networking are based on closed control loop allowing self-adapting (self-configuring and self-optimizing) of the network equipment according to the events which arise on their environment. Knowledge plane is one approach, very emphasis these last years by researchers, which suggests the use of the cognitive systems (machine learning and the reasoning) to realize closed control loop. However, although the major autonomic architectures integrate machine learning modules as functional block, few researches are really interested in the contents of these blocks. It is in this context that we made a study on the potential contribution machine learning and proposed a method of distributed and collaborative machine learning. We propose a formalization self-adapting problem in term of learning configuration strategies (state-actions) problem. This formalization allows us to define a strategies machine learning method for self-adapting which is based on the history observed transitions and uses inductive logic programming to discover new strategies from those already discovered. We defined, also a knowledge sharing algorithm which makes network components collaborate to improve learning process. Finally, we tested our approach in DiffServ context and showed its transposition on multimedia streaming in 802.11 wireless networks
Piat, Guilhem Xavier. "Incorporating expert knowledge in deep neural networks for domain adaptation in natural language processing". Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASG087.
Texto completoCurrent state-of-the-art Language Models (LMs) are able to converse, summarize, translate, solve novel problems, reason, and use abstract concepts at a near-human level. However, to achieve such abilities, and in particular to acquire ``common sense'' and domain-specific knowledge, they require vast amounts of text, which are not available in all languages or domains. Additionally, their computational requirements are out of reach for most organizations, limiting their potential for specificity and their applicability in the context of sensitive data.Knowledge Graphs (KGs) are sources of structured knowledge which associate linguistic concepts through semantic relations. These graphs are sources of high quality knowledge which pre-exist in a variety of otherwise low-resource domains, and are denser in information than typical text. By allowing LMs to leverage these information structures, we could remove the burden of memorizing facts from LMs, reducing the amount of text and computation required to train them and allowing us to update their knowledge with little to no additional training by updating the KGs, therefore broadening their scope of applicability and making them more democratizable.Various approaches have succeeded in improving Transformer-based LMs using KGs. However, most of them unrealistically assume the problem of Entity Linking (EL), i.e. determining which KG concepts are present in the text, is solved upstream. This thesis covers the limitations of handling EL as an upstream task. It goes on to examine the possibility of learning EL jointly with language modeling, and finds that while this is a viable strategy, it does little to decrease the LM's reliance on in-domain text. Lastly, this thesis covers the strategy of using KGs to generate text in order to leverage LMs' linguistic abilities and finds that even naïve implementations of this approach can result in measurable improvements on in-domain language processing
Bouraoui, Basma. "Architectures cross-layer PHY/MAC pour réduire l'effet de blocage de réception dans les réseaux véhiculaires ad-hoc". Thesis, Valenciennes, 2017. http://www.theses.fr/2017VALE0006/document.
Texto completoThe MAC protocol IEEE 802.11p, dedicated to vehicular ad-hoc networks VANETs, prohibits simultaneous transmissions in the same detection area, in order to avoid interference between neighboring vehicles. This prohibition causes a temporary data reception blocking, which reduces the network throughput. To reduce this adverse impact, we propose in this thesis a cross-layer design PHY/MAC based on a transmit antennas selection algorithm jointly with a dedicated MAC protocol. This design allows receivers to select the best combination of transmit antennas to improve the throughput of each V2V link. The algorithm is presented with a multi-user detection method, which cancels neighbor’s interference and allows vehicles to transmit data simultaneously. The associated MAC protocol ensures the coordination between vehicles during the simultaneous transmission period. The simulation results show a significant network throughput improvement compared to the conventional case. However, this improvement is less important in dense VANETs. For this purpose, we propose to join a cross-layer design PHY/MAC based on a transmit power adaptation algorithm. This design allows transmitters to choice the adequate power level based on corresponding receivers density. The simulation results show that this solution allows more vehicles to communicate simultaneously and thus improves the network throughput, in particular in dense VANETs
Buhrig, Aurélien. "Optimisation de la consommation des noeuds de réseaux de capteurs sans fil". Phd thesis, Grenoble INPG, 2008. http://tel.archives-ouvertes.fr/tel-00319073.
Texto completoBuhrig, Aurélien. "Optimisation de la consommation des noeuds de réseaux de capteurs sans fil". Phd thesis, Grenoble INPG, 2008. http://www.theses.fr/2008INPG0027.
Texto completoWireless sensor networks pose many design challenges. They must collect information coming from the environment, treat gathered data, receive and forward it with a lifetime that must reach tens of years depending on the applications without any external intervention. It is thus necessary to optimize the energy consumption at every design level. This work proposes to reduce the energy consumption of the digital part of a network node thanks to the use of asynchronous logic and synchronous communications between software and hardware. In addition, it is important to dynamically manage consumption by scaling the supply voltages. This aspect is treated on the whole system, at the material level, thanks to the specification of a dedicated coprocessor, at the algorithmic level and real-time software, and at the communication interfaces between the software and the hardware
Majed, Najmeddine. "Measuring and improving the quality of experience of mobile voice over IP". Thesis, Ecole nationale supérieure Mines-Télécom Atlantique Bretagne Pays de la Loire, 2018. http://www.theses.fr/2018IMTA0099/document.
Texto completoFourth-generation mobile networks, based on the Long Term Evolution (LTE) standard, are all- IP networks. Thus, mobile telephony providers are facing new types of quality degradations related to the voice packet transport over IP network such as delay, jitter and packet loss. These factors can heavily degrade voice communications quality. The real-time constraint of such services makes them highly sensitive to delay and loss. Network providers have implemented several network optimizations for voice transport to enhance perceived quality. However, the proprietary quality management algorithms implemented in terminals are left unspecified in the standards. In this context, we are interested in media adaptation mechanisms integrated in terminals to enhance the overall Quality of Experience (QoE). In particular, we experimentally evaluate Voice over LTE (VoLTE) QoE metrics such as delay and Mean Opinion Score (MOS) sing a standardized test method. We propose some enhancements to the actual test method and discuss how this method can be extended to evaluate de-jitter buffer performance. We also experimentally evaluate WebRTC voice quality in different radio conditions using a realLTE test network. We evaluate the impact of jitter buffer and bit rate variations on the measured quality. To enhance voice codec robustness against packet loss, we propose a simple application layer redundancy. We implemented it for the Enhanced Voice Service (EVS) codec and evaluate it. Finally, we propose a signaling protocol that allows sending redundancy requests during a call to dynamically activate or deactivate the redundancy mechanism
Tchuente, Dieudonné. "Modélisation et dérivation de profils utilisateurs à partir de réseaux sociaux : approche à partir de communautés de réseaux k-égocentriques". Toulouse 3, 2013. http://thesesups.ups-tlse.fr/1972/.
Texto completoIn most systems that require user modeling to adapt information to each user's specific need, a user is usually represented by a user profile in the form of his interests. These interests are learnt and enriched over time from users interactions with the system. By the evolving nature of user's interests, the user's profile can never be considered fully known by a system. This partial knowledge of the user profile at any time t significantly reduces the performance of adaptive systems, when the user's profile contains no or only some information. This drawback is particularly most recurrent for new users in a system (time t = 0, also called cold start problem) and for less active users. To address this problem, several studies have explored data sources other than those produced by the user in the system: activities of users with similar behavior (e. G. Collaborative filtering techniques) or data generated by the user in other systems (e. G. , multi-application user's profiles, multiple identities management systems). By the recent advent of Social Web and the explosion of online social networks sites, social networks are more and more studied as an external data source that can be used to enrich users' profiles. This has led to the emergence of new social information filtering techniques (e. G. Social information retrieval, social recommender systems). Current studies on social information filtering show that this new research field is very promising. However, much remains to be done to complement and enhance these studies. We particularly address two drawbacks: (i) each existing social information filtering approach is specific in its field scope (and associated mechanisms), (ii) these approaches unilaterally use profiles of individuals around the user in the social network to improve traditional information filtering systems. To overcome these drawbacks in this thesis, we aim at defining a generic social model of users' profiles that can be reusable in many application domains and for several social information filtering mechanisms, and proposing optimal techniques for enriching user's profile from the user's social network. We rely on existing studies in social sciences to propose a communities (rather than individuals) based approach for using individuals around the user in a specific part of his social network, to derive his social profile (profile that contains user's interest derived from his social network). The significant part of the user's social network used in our studies is composed of individuals located at a maximum distance k (in the entire social network) from the user, and relationships between these individuals (k-egocentric network). Two evaluations of the proposed approach based on communities in k-egocentric networks have been conducted in the online social network Facebook and the co-authors network DBLP. They allow us to demonstrate the relevance of the proposal with respect to existing individual based approaches, and the impact of structural measures such as the centrality of communities (degree or proximity) or user's k-egocentric network density, on the quality of results. Our approach opens up many opportunities for future studies in social information filtering and many application domains as well as on the Web (e. G. Personalization of search engines, recommender systems in e-commerce, adaptive systems in e-Learning environment) or in Intranets business systems (e. G. Behavioral analysis in networks of subscribers telecom customers, detection of abnormal behavior network bank customers, etc. )
Benamar, Alexandra. "Évaluation et adaptation de plongements lexicaux au domaine à travers l'exploitation de connaissances syntaxiques et sémantiques". Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASG035.
Texto completoWord embeddings have established themselves as the most popular representation in NLP. To achieve good performance, they require training on large data sets mainly from the general domain and are frequently finetuned for specialty data. However, finetuning is a resource-intensive practice and its effectiveness is controversial.In this thesis, we evaluate the use of word embedding models on specialty corpora and show that proximity between the vocabularies of the training and application data plays a major role in the representation of out-of-vocabulary terms. We observe that this is mainly due to the initial tokenization of words and propose a measure to compute the impact of the tokenization of words on their representation. To solve this problem, we propose two methods for injecting linguistic knowledge into representations generated by Transformers: one at the data level and the other at the model level. Our research demonstrates that adding syntactic and semantic context can improve the application of self-supervised models to specialty domains, both for vocabulary representation and for NLP tasks.The proposed methods can be used for any language with linguistic information or external knowledge available. The code used for the experiments has been published to facilitate reproducibility and measures have been taken to limit the environmental impact by reducing the number of experiments
Da, Cunha Violette. "Réseaux de régulation et éléments intégratifs et conjugatifs de la famille TnGBS dans l'adaptation de Streptococcus agalactiae". Phd thesis, Université Pierre et Marie Curie - Paris VI, 2012. http://tel.archives-ouvertes.fr/tel-00828193.
Texto completoAbdeddaim, Mohamed Nazim. "Analyse des performances d'un réseau de capteurs exploitant le standard IEEE 802.15.4". Thesis, Grenoble, 2012. http://www.theses.fr/2012GRENM067/document.
Texto completoAn increasing interest has been observed in Wirless Sensor Network that can be explained by wide range of WSN applications as well as by the challenges involving the constraints of this type of networks. The IEEE 802.15.4 standard has been proposed with the objective of standardizing the physical and MAC layers. In this work, we have firstly proposed an alternative multichannel scheme for the IEEE 802.15.4 standard. It is able to solve the problem of superframe collisions based on a cluster-tree topology approach in which each cluster uses a different channel allowing the reduction of the interference and increased network capacity. We have also difined a novel mechanism for topology construction, channel allocation, and superframe scheduling. Secondly, we have analyzed the impact of different parameters on the medium access control. We have shown the weakness of the medium access method proposed in the standard. For instance, we have observed a decrease in performance when the standard is applied in large networks. Motivated by the analysis and its results we have then proposed auto-adaptive mechanisms for the medium access control. They allow to dynamically adapt the size of the contention window according to the observed traffic conditions. Each coordinator computes the optimal values to avoid deafness. The proposed mechanisms rapidly converge even in the case of bursty traffic
Vergotte, Grégoire. "Adaptability and adaptation to a sensorimotor task : from functional significance of fractal properties to brain networks dynamics". Thesis, Montpellier, 2018. http://www.theses.fr/2018MONT4004/document.
Texto completoThe study of fractal properties in biological time series is of increasing interest. Nevertheless, the literature highlights an ambiguity on the causal explanation of the presence of these time series which does not make it possible to distinguish between the effective adaptation made by a subject or his overall adaptability capacities. The aim of this dissertation is to decorrelate these two notions, notably by linking the behavioral level to the cerebral level. Our first study allowed to highlight that the mono-fractal properties could reflect the adaptability of the subjects whereas the multifractal properties would be related to the effective adaptation carried out during the task. The second study showed a correlation between the multifractal properties and the number of brain networks implemented during the task, reflecting the effective adaptation to the experimental constraints imposed. The results of this work have allowed us to better understand the functional meaning of fractal analyzes in terms of effective adaptation and adaptability