Taoufik Yeferny

Ces deux dernières décennies les systèmes P2P de partage de fichiers sont devenus très populaires grâce aux accès à des ressources diverses, distribuées sur Internet. Parallèlement à l'évolution de cette catégorie de systèmes, les dispositifs mobiles (téléphones cellulaires, PDA et autres appareils portatifs) ont eu un grand succès sur le marché. Équipés d'une technologie de communication sans fil (Bluetooth, et Wifi), ils peuvent communiquer sans nécessiter une infrastructure particulière en utilisant un réseau mobile adhoc (Mobile Adhoc NETwork -MANET). De la même manière, les systèmes P2P peuvent être aussi déployés sur ce type de réseau et deviennent des systèmes P2P mobiles (Mobile 2P systems). Dans le cadre de cette thèse, nous nous intéressons essentiellement à la recherche d'information dans les systèmes P2P et plus précisément au problème de routage de requêtes. La première partie de la thèse, s'est focalisée sur le routage de requêtes dans les systèmes P2P sur Internet. Nous avons proposé (i) un modèle de routage sémantique basé sur l'historique des requêtes. Ce modèle est ensuite instancié pour définir une nouvelle méthode de routage par apprentissage. Pour pallier le problème de démarrage à froid, (ii) nous avons proposé une méthode prédictive de l'intention de l'utilisateur qui construit une base de connaissances à priori pour chaque pair. Enfin, (iii) nous avons proposé une méthode de routage hybride pour traiter le problème d'échec de sélection. Cette méthode est basée sur l'historique des requêtes et le regroupement de pairs dans des groupes sémantiques. La deuxième partie de la thèse, s'est focalisée sur le routage de requêtes dans les systèmes P2P mobiles. L'apparition des MANETs, a soulevé de nouveaux challenges de routage. Ces réseaux souffrent de plusieurs contraintes liées aux supports de transmission ou bien aux dispositifs mobiles. Dans ce cadre, nous avons proposé une méthode de routage pour les systèmes P2P non structurés mobiles basée sur le contexte de l'utilisateur. D'un point de vue technique, toutes ces propositions ont été développées, validées et évaluées grâce aux simulateurs PeerSim et NS2

Peer-to-peer (P2P) systems have offered users an efficient way to share various resources and access diverse services over the Internet. In unstructured P2P systems, resource storage and indexation are fully distributed among participating peers. Therefore, locating peers sharing pertinent resources for a specific user query is a challenging issue. In fact, effective query routing requires smart decisions to select a certain number of peers with respect to their relevance for the query instead of choosing them at random. In this respect, we introduce here a new query-oriented approach, called the reinforcement learning-based query routing approach (RLQR). The main goal of RLQR is to reach high retrieval effectiveness as well as a lower search cost by reducing the number of exchanged messages and contacted peers. To achieve this, the RLQR relies on information gathered from previously sent queries to identify relevant peers for forthcoming queries. Indeed, we formulate the query routing issue as the reinforcement learning problem and introduce a fully distributed approach for addressing it. In addition, RLQR addresses the well-known cold-start issue during the training stage, which allows it to improve its retrieval effectiveness and search cost continuously, and, therefore, goes quickly through the cold-start phase. Performed simulations demonstrate that RLQR outperforms pioneering query routing approaches in terms of retrieval effectiveness and communications cost.

P2P networking has grasped an increasing interest worldwide among both researchers and computer networking professionals. As a witness, several P2P applications mainly used for file sharing over the Internet have been proposed. Considering the great success of mobile devices in recent years, P2P applications have also been deployed over mobile networks such as mobile ad-hoc networks (MANETs). However, the mismatch between the P2P overlay and the MANET underlay topologies makes the resources lookup mechanism in mobile P2P applications very difficult. Therefore, this downside is the main hindrance to the deployment of such applications over MANETs. To overcome the mismatch issue, we propose in this paper RLSM-P2P a cross-layer resource lookup scheme for Mobile P2P applications. The main thrust of RLSM-P2P consists of building an efficient unstructured P2P overlay that closely matches the underlay physical network and swiftly adapts to its volatility and dynamicity by considering different MANET constraints. Furthermore, RLSM-P2P relies on a query learning resource lookup mechanism for locating pertinent resources to user queries. The performed experiments show that RLSM-P2P outperforms its competitors in terms of effectiveness and efficiency.

Summary: A Vehicle AdHoc Network is mainly composed of mobile vehicles and fixed Road Site Units (RSUs). The latter is usually very expensive to deploy and has a crucial role in maintaining the network connectivity. Therefore, the design of an efficient RSU deployment strategy that enables a high coverage ratio and a lower deployment cost has been of paramount importance. In this respect, we introduce in this paper a new spatiotemporal coverage strategy for nonsafety Vehicle AdHoc Network applications like driving assistance and business promotion, called Minimal Mobility Patterns Coverage (MPC). The main thrust of MPC is to (1) depict the mobility patterns of moving vehicles from their trace files and then (2) compute the adequate RSU locations in order to cover the extracted mobility patterns by the minimal possible number of RSUs. To this end, we firstly provide a new method to depict the mobility patterns of vehicles by mining the correlations between the kept track connections of vehicle trajectories versus crossed junctions. Secondly, we introduce a new way to compute the adequate RSU locations through the instantiation of the well‐known problem of extracting minimal transversals of a hypergraph. Experimental results show that our RSUs deployment strategy performs better than baseline strategies.

The main objective of the VANET networks is to improve road safety as well as transport efficiency through the use of communications technology and the emergence of wireless devices at low cost. Thus, the design of an efficient dissemination protocol, that informs vehicles about interesting safety and nonsafety events, is of paramount importance. The thriving challenge would be to maximize the delivery ratio by avoiding as far as possible the broadcast storm problem. A scrutiny of the literature wealthy number of approaches highlights that all of them fail to fulfill with the "sine qua non" requirements that we introduce. In this paper and to palliate this shortage, we introduce a new infrastructure-less geocast approach, called DPMS (data Dissemination Protocol based on Map Splitting). The main thrust of DPMS stands in reaching a high delivery ratio as well as a high geocast precision by sending messages only to vehicles in the Zone of Relevance (ZOR) with a minimum overhead cost. Carried out experiments showed that DPMS outperforms its competitors in terms of effectiveness and efficiency.