Doctorat LMD RSD
Permanent URI for this collectionhttps://dspace.univ-tlemcen.dz/handle/112/10248
Browse
Recent Submissions
Item L’intelligence Artificielle au Service de l’E-santé : Applications Pour le Diagnostic des Maladies de la Peau(University of Tlemcen, 2025-02-27) M'Hamedii, MohammedLes affections cutanées, dont les cancers de la peau, représentent un défi important pour la santé publique. L’apparence souvent trompeuse des lésions souligne l’importance d’un diagnostic précis par un dermatologue. Parmi les cancers de la peau, le mélanome malin se distingue par sa gravité et sa capacité à se métastaser rapidement. Bien qu’il soit moins fréquent que d’autres types de cancer cutané, comme le carcinome basocellulaire et le carcinome spinocellulaire, il nécessite une attention particulière en raison de son pronostic. La dermoscopie est un outil essentiel pour les dermatologues car elle permet de détecter les mélanomes à un stade précoce, ce qui est crucial pour un traitement efficace. La combinaison de la dermoscopie et de l’intelligence artificielle offre de nouvelles perspectives pour un diagnostic plus précis et plus rapide du mélanome. Les algorithmes d’apprentissage profond, entraînés sur des datasets d’images de haute qualité, peuvent aider les dermatologues à détecter les mélanomes à un stade précoce, améliorant ainsi les chances de guérison des patients. Cette thèse présente le développement d’un système d’aide au diagnostic du cancer de la peau basé sur l’apprentissage profond, visant à classifier précisément le Mélanome malin. L’étude explore diverses architectures de réseaux neuronaux convolutifs (CNN), en intégrant l’apprentissage par transfert pour tirer parti des modèles pré-entraînés, l’augmentation de données pour enrichir les ensembles d’entraînement, et des architectures hybrides CNN-LSTM pour améliorer les performances du modèle. L'évaluation expérimentale a démontré que l’architecture MobileNetV2-LSTM offre les meilleures performances en termes d’exactitude, sensibilité et spécificité, surpassant les approches existantes dans la littérature. L’objectif est d’améliorer la détection précoce du mélanome grâce à des méthodes computationnelles avancées, ce qui pourrait potentiellement augmenter les taux de survie des patients.Item Sélection des services web avec prise en compte de la qualité de service dynamique et incertaine(University of Tlemcen, 2024-04-20) Etchiali, AbdelhakLa technologie des services web constitue une implémentation idéale du paradigme du calcul orienté services (SOC). Étant donné que l’objectif principal du SOC est d’assurer l’interopérabilité des applications et la création de compositions d’applications (ou de services) avec valeurs ajoutées, il conviendra de concevoir et de mettre en oeuvre des modèles permettant de combiner des services web individuels dans des workflows satisfaisants des critères de performance objectifs. Il convient de noter que les services web courants sont caractérisés par différents attributs de QoS qui jouent un rôle majeur dans la spécification des compositions de services désirés. Il est utile de souligner que les attributs de QoS dépendent largement des fluctuations de l’environnement (par exemple, la surcharge des réseaux, ou la fluctuation des coûts en raison des saisons ou des événements socioculturels) et par conséquent, leur incertitude créera des difficultés supplémentaires dans la modélisation mathématique du problème de composition. Dans cette thèse, nous adressons la composition des services avec incertitude de QoS en proposant deux contributions principales, toutes les deux exploitent une recherche locale et globale pour alléger la complexité temporelle du problème. La première contribution exploite l’heuristique des intervalles majoritaires pour effectuer la recherche locale, en outre, la recherche globale est effectuée à l’aide d’une recherche exhaustive qui exploite les contraintes globales. Dans la deuxième contribution, nous adoptons une version discrète de la méta-heuristique de l’algorithme des chauves-souris (bat algorithm) en plus d’un ensemble d’heuristiques (telles que la dominance floue et la dominance stochastique d’ordre zéro) pour effectuer à la fois la recherche locale et globale. Les résultats obtenus confirment l’efficacité de nos contributions, et en particulier, les performances étaient satisfaisantes pour les workflows qui ont une taille variant entre 2 et 10 composants.Item Privacy dans l’IdO basée sur une approche cryptographique et non cryptographique(University of Tlemcen, 2024-04-07) Sebbah, AbderrezzakL'Internet des Objets (IdO) a émergé comme un secteur dynamique de l'internet, capturant un intérêt significatif ces dernières années et ayant un impact transformationnel sur la société. Ce domaine, regroupant une gamme variée d'objets connectés de tailles diverses, offre une multitude d'applications dans des domaines allant de la domotique à l'agriculture, en passant par la sécurité, les transports et la santé. L'IdO vise à faciliter l'interaction entre les objets agissant comme des capteurs et des actionneurs, permettant ainsi le contrôle à distance des appareils intelligents via des connexions ouvertes. Néanmoins, cette ouverture expose les données sensibles à un large éventail de risques, les rendant vulnérables à diverses formes d'attaques. La sécurisation de ces réseaux représente un défi majeur, en particulier dans des environnements caractérisés par des ressources limitées et une diversité importante. Pour répondre à ces défis, nous proposons trois mécanismes de sécurité robustes pour les systèmes IdO, conçus pour contrer les menaces émanant des connexions ouvertes. L'évaluation de ces solutions, réalisée à l'aide de la logique de Burrows-Abadi-Needham (BAN) et de l'outil de validation automatisée AVISPA, démontre leur fiabilité, leur efficacité et leur adaptation aux spécificités des réseaux IdO, comparativement à d'autres techniques récentes similaires.Item Hardy-Sobolev equations in p-Laplacian on compact Riemannian manifolds.(University of Tlemcen, 2024-06-19) Ghomari, Mohammed TewfikIn this thesis we study, on compact Riemannian manifolds, a quasi-linear elliptic equation in p-Laplacian operator containing a Hardy term and a critical Sobolev exponent. We first show that Palais-Smale sequences of our equation are submitted to the well known Struwe decomposition formulas. In a second part, we prove some existence results relying on the decomposition results.Item Optimisation des systèmes de e-santé à base de l’internet des objets(University of Tlemcen, 2023-05-20) Zerga, HideyatItem Étude de solutions Cloud pour les communications véhiculaires(University of Tlemcen, 2023-10-28) Gaouar, NihalItem Spark au service d’ETL pour la gestion des données RDF streaming(University of Tlemcen, 2023-11-04) Gueddoudj, El YazidItem gestion efficace de Big Data dans le contexte Spatial et RDF(University of Tlemcen, 2023-12-07) Yousfi, Houssam EddineItem Sécurité des données et protection de la vie privée des usagers dans les réseaux Fog véhiculaires(University of Tlemcen, 2023-11-25) Benadla, SarraItem Privacy preserving IoT-based healthcare data using fog-to-cloud computing(University of Tlemcen, 2023-07-06) Saidi, HafidaOver the past few decades, the world has become more connected with the wide adoption of Internet of Things (IoT), cloud computing, and fog computing. These technologies are the driving force to collect, process, and store medical data. However, the privacy and security of health data represent major challenges. For this purpose, to enhance the security and benefit from the advantages of cloud and fog computing, a hierarchical Fog-To-Cloud (F2C) computing system was introduced which integrates the fog and the cloud in a single model. In this thesis, we provide a comprehensive state-of-the-art that deals with the aforementioned problem in the context of IoT, F2C, and e-health. Then, we propose two contributions using several technologies. As the patient’s medical data are accessible by users who have diverse privileges, we have adopted a decentralized access control system using blockchain and Self-Sovereign Identity (SSI) for privacy-preserving data. Hence, our proposed approach focuses on smart contract to conduct Role-Based Access Control policies (RBAC) and adopts the implementation of Decentralized IDentifiers (DID) and Verifiable Credentials (VC) to describe advanced access control techniques for emergency cases. Experimental results based on privacy-preserving medical records demonstrate that our proposed solution ensures a high level of security, protect data privacy, empower patients with mechanisms to preserve control over their personal information, and allow them to self-grant access rights to their medical data.Item Privacy dans l’IdO basée sur une approche cryptographique et non cryptographique(University of Tlemcen, 2024-04-07) Sebbah, AbderrezzakL'Internet des Objets (IdO) a émergé comme un secteur dynamique de l'internet, capturant un intérêt significatif ces dernières années et ayant un impact transformationnel sur la société. Ce domaine, regroupant une gamme variée d'objets connectés de tailles diverses, offre une multitude d'applications dans des domaines allant de la domotique à l'agriculture, en passant par la sécurité, les transports et la santé. L'IdO vise à faciliter l'interaction entre les objets agissant comme des capteurs et des actionneurs, permettant ainsi le contrôle à distance des appareils intelligents via des connexions ouvertes. Néanmoins, cette ouverture expose les données sensibles à un large éventail de risques, les rendant vulnérables à diverses formes d'attaques. La sécurisation de ces réseaux représente un défi majeur, en particulier dans des environnements caractérisés par des ressources limitées et une diversité importante. Pour répondre à ces défis, nous proposons trois mécanismes de sécurité robustes pour les systèmes IdO, conçus pour contrer les menaces émanant des connexions ouvertes. L'évaluation de ces solutions, réalisée à l'aide de la logique de Burrows-Abadi-Needham (BAN) et de l'outil de validation automatisée AVISPA, démontre leur fiabilité, leur efficacité et leur adaptation aux spécificités des réseaux IdO, comparativement à d'autres techniques récentes similaires.Item Item Développement d'un mécanisme de détection d'intrusion dans les réseaux de capteurs corporels sans fil (WBANs).(21-11-2022, 2022-05-25) Achour, M’HammedIn this work, we chose the IEEE 802.15.4 standard as an enabling technology for wireless body area networks. To secure these networks, we examined the beacon mode of the standard from a security perspective to find threats that target network availability. As a result, we have introduced a new attack that exploits the behavior of the standard in the case of periodic traffic to disturb it using as few resources as possible. To complete our work’s objective, we have proposed a countermeasure whose goal is to reduce the damage of this attack and to take back the majority of network resources in the case of a simple attacker. Moreover, we have proposed an algorithm that estimates the natural packet error ratio to detect anomalies in the periodic traffic of IEEE 802.15.4 based networks. To achieve this goal, we had to transform the periods of the nodes in order to capture the seasonality in the traffic. The algorithm did well in terms of false positive and detection ratios.Item Conceptualisation des bonnes pratiques au sein d'une communauté de pratique.(17-11-2022, 2022-06-28) Hamza-Cherif Épouse Rahmoun, SouaadSince the advent of the social and semantic web, in recent years new tools and sharing sites such as Meta, Twitter, WikiHow, etc. have emerged, making the web a universal collection of knowledge, where users geographically form communities of practice (CoP) online, these CoPs are originally a concept of sociology but find their full development in the current web where users share and exchange their know-how in different fields in the form of procedural knowledge ( PK) called good practices. These good practices are defined by a set of successive steps taken to achieve an objective. Conceptualizing this procedural knowledge has become a major challenge in several fields (information retrieval, intelligent applications, robotics...), knowledge extraction from data base (KDD) is the field that is evolving to offer solutions. KDD combines different methods of learning and knowledge representation in order to find solutions to explore unstructured data in order to facilitate their exploitation and in this context several works have focused on the exploration of procedural knowledge in different purposes, sometimes to create a knowledge base or to identify instructions from procedural knowledge. Most of this work is in the field of natural language processing, the goal we pursue is another, in this thesis we present a new approach to extract and conceptualize good practices from the web, and extract the best practice for a given query. The proposed approach takes place in two phases: in the first one extracts good practices from the web using a web scrapping method, after we represent them by oriented data graphs. In the second phase, we extract the best practice for a given query by applying the techniques of machine learning and text summarization on graphs. This phase takes place in three steps: (1) search for practices similar to the user’s query, here we use the word embedding model to identify sentences similar to the goal sought by the user; (2) Grouping and fusion of similar steps, where we use unsupervised learning (DBScan) and text summarization (PageRank) techniques to group semantically close nodes that we merge in the same step; (3) Extraction of the best practice that is identified by the path of the graph traversing the most important steps to reach the objective, this importance is calculated by measures of centrality of the graphs which quantify the importance of the nodes in a graph oriented by the number of their incoming and outgoing arc. The results obtained demonstrated the superiority of our approach for: (1) capturing practices similar to the goal sought by the user, and this by optimizing the execution time, (2) extracting the best practices for queries compared to a search engine from a real data set.Item MULTI-CLASSIFIEURS DES IMAGES SATELLITAIRES.(16-11-2022, 2022-06-18) Chaouche Ép. Stambouli, Ramdane LamiaIn remote sensing, clustering, also called unsupervised classification, is an important task that aims to partition a given image in a multispectral space into a number of spectral classes (clusters), when in situ information is not available. Among the many existing clustering algorithms, the most commonly used are K-means, ISODATA, FCM (Fuzzy C-Means), SOM (Self Organizing Map) and more recently K-Harmonic Means. However, with the increase in the amount of remotely sensed data and its heterogeneity, it becomes difficult to obtain relevant clustering results using a single clustering algorithm. Moreover, each algorithm mentioned above requires a number of parameters and the most important of them is the number of clusters, which the user has to define a priori. To cope with these shortcomings, the Multiple Classifier System (MCS) is also known as ensemble clustering , is the consensus of different clustering algorithms can provide the best partition with high accuracy and consequently overcome limitations of traditional approaches based on single classifiers. The MCS involves two stages : the partitions generation and the partitions combination. In this thesis, we investigate the potential advantages of this technique in the unsupervised land cover classification by using various kinds of data : Synthetic data, composite data and remotely sensed data. The first stage of the MCS is assumed by four clustering algorithms, the well-known k-means algorithm, the k-harmonic means algorithm (KHM), Bisecting K-means (BKM) and the self-organizing map (SOM). The best clustering is obtained according to WB index. The relabeling and the voting methods are used in the second stage. Experimental results obtained by the MCS outperform the results of the individual clustering.Item Reliable data dissemination in Vehicular Ad-hoc Networks.(22-06-2022, 2022-07-05) Lazhar, KhamerIn recent years, the decentralized wireless Vehicular Ad hoc Networks (VANETs) have emerged as a key technology for Intelligent Transportation Systems (ITS). Efficient and reliable multi-hop broadcast protocols are essential to support various services in VANETs such as road safety, traffic efficiency, entertainment, and advertising. The multi-hop broadcasting protocol intends to deliver data to a set of vehicles inside a region of interest (RoI). Data broadcasting requires different levels of quality of service, which can be specified based on the type of data included in the message. For instance, accident notification involves low latency and high packets delivery ratio, whereas congested road notification is tolerant to both end-to-end delay (around a couple of seconds) and packets delivery ratio without exposing road users to a dangerous situation. Besides, the delay, jitter, and packet loss ratio associated with data video broadcasting should not exceed strict thresholds for an acceptable quality of experience. Our main aim in this thesis is to design reliable multi-hop broadcast protocols for delay-tolerant applications and video streaming in urban VANETs.Item Estimation de la QoS dans les services Web par apprentissage profond.(22-06-2022, 2022-03-12) Smahi, Mohammed IsmailIn this thesis, we propose a deep learning-based approach wich combines a matrix factorization model based on a deep auto-encoder (DAE) and a clustering technique. Three variants of the auto-encoder design have been used. The first one is composed of a single hidden layer that represents the vector of latent factors of users and/or services. A second architecture considers several hidden layers. A third model consists of a combination of a deep auto-encoder model and a generative adversarial network. Other problems underlying the estimation of missing QoS values were addressed in this work. The first one is related to the vulnerability of prediction systems to the data sparsity problem. To deal with this issue our proposal consists of in using a clustering algorithm based on Kohonen’s self-organising maps, where the initialization is done using location attributes. The second one that we have dealt with is the cold start problem, which occurs when adding new users/services to the prediction system. The latter one is globally managed by exploiting a spatial features as well. The conducted experiments show that our proposals can provide better performances in terms of QoS prediction, and consequently provide more guidance for users in their choice of preferred services than existing methods do. The QoS parameters on which we relied on to carry out our various experiments are response time and throughput. However, the proposed QoS prediction algorithms can be applied to other QoS factors.Item Détection et communication coopérative pour le déploiement des réseaux de capteurs.(22-06-2022, 2022-03-05) Benaissa, Bedr-EddineOperating a sensor network raises many issues at several algorithmic levels: localization, deployment, data collection, coverage, and reduction of battery power consumption to optimize network lifetime. This last point has been of particular interest to the researchers. In such a network, and especially with a single-hop deployment policy, sensor measurements contain a lot of redundancy, either in the measurement dimensions of a single sensor, or between the measurement dimensions of different sensors due to spatial correlation or in the temporal dimension of measurements. The goal is to study detection and cooperation to determine conditions that will help to better position sensors in a given deployment area, while guaranteeing certain constraints related to this type of network, such as the cost of deployment and the network's lifetime. Two approaches have been proposed. The first one proposes to minimize the complexity in terms of communication and computation by relying on an aggregation and consensus system to reduce the spatial and temporal dimension of the captured data and consequently the number of deployed sensors. The results show a visible performance compared to the standard transmission method on the open platform of the COOJA / Contiki simulator allowing to simulate wireless sensor network connections and to interact with them. The second contribution minimizes the transmission frequencies of the measured data to the base station by categorizing the captured data into predefined, prenumbered classes, which we will call "confidence intervals". In this way, each captured value will be classified into a class and only its number will be sent to the base station, if (and only if) a class change with respect to the previous value is observed. The results show that interval-based data collection significantly reduces the energy of the motes' sensors. Thus, in a wireless sensor network, optimal deployment is accentuated by good data transmission management to the sink.Item Routage multi-chemin dans les réseaux de capteurs multimédias sans fil.(19-06-2022, 2021-07-10) Chikh, AsmaMultipath geographic routing is an effective strategy for transmitting data flows in wireless multimedia sensor networks (WMSNs) since it enables to meet the Quality of Service (QoS) requirements in this type of networks. The first part of this thesis proposes two states of the art concerning multipath routing solutions developed in the literature for WMSNs. First, a new classification of these protocols was proposed based on the purpose of routing. Second, the multipath geographic routing solutions have been classified according to the original solutions and their derivatives. The second part of this thesis consists to improve, develop and design new multipath geographic routing solutions in WMSNs. The first protocol proposed is a multi-path version of the GPSR protocol called MMGPSR. Next, we designed an AGEM protocol extension where a triangular link quality metric is added to select the next hop called TLQM-AGEM. Third, a new multipath geographic routing protocol is proposed, called LQLB-MGR, this routing solution consists of two phases where the first is intended for building the set of paths while the second is for load balancing between the paths selected during the first phase. Finally, another new multi-path geographic protocol is proposed based on the two greedy routing approaches called GCGM, this protocol also proposed weighted constants for all metrics used for the selection of paths. The simulation results illustrated the benefits of our routing solutions in terms of several performance criteria such as the network lifetime, the packet delivery rate (PDR) and the power consumption.Item Safety-Oriented Identity and Location Preservation in Internet of Vehicles.(19-06-2022, 2021-06-08) Babaghayou, MessaoudThis thesis deals with the problem of identity and location privacy in the context of Internet of Vehicles (IoV) while making road-safety into consideration. This problematic emerged with the advent of different safety-achieving techniques provided by IoV applications. There exist many techniques that cope with the identity and location privacy problem but while sacrificing safety. In our thesis, we focus on the solutions that are based on the pseudonymity concept and many techniques related to this category were proposed. With this said, we provide a comprehensive survey that deals with the aforementioned problem. then, we propose three techniques that ensure high level of location privacy while considering road-safety as an objective. The obtained results show that road-safety can still be achieved in conjunction with location privacy while using our techniques.