Tese de mestrado. Redes e Serviços de Comunicação. Faculdade de Engenharia. Universidade do Porto. 2009

Delay Tolerant Network (DTN) is emerging as a solution for supporting data transfer in intermittently connected networks. In DTN, to cope with long disconnections, messages are buffered for a long period of time. Thus, according to the queue management the performance can be affected significantly. Power is also a scarce resource in DTN. Energy can be saved by putting mobile nodes into sleep during long delayed connections. In this thesis, a Medium Access Control (MAC) protocol that supports adaptive sleep scheduling of a mobile node is proposed. Based on the MAC layer operation, an adaptive power management framework is developed. The framework considers power saving and buffer management together in order to minimize power consumption while minimizing the performance degradation of buffer management for the mobile node. Variations of the performance of a traffic source node which are affected by diverse network parameters are also investigated.

Delay Tolerant Network (DTN) is emerging as a solution for supporting data transfer in intermittently connected networks. In DTN, to cope with long disconnections, messages are buffered for a long period of time. Thus, according to the queue management the performance can be affected significantly. Power is also a scarce resource in DTN. Energy can be saved by putting mobile nodes into sleep during long delayed connections. In this thesis, a Medium Access Control (MAC) protocol that supports adaptive sleep scheduling of a mobile node is proposed. Based on the MAC layer operation, an adaptive power management framework is developed. The framework considers power saving and buffer management together in order to minimize power consumption while minimizing the performance degradation of buffer management for the mobile node. Variations of the performance of a traffic source node which are affected by diverse network parameters are also investigated.

There has been much research in recent years within the general field of mobile ad hoc networks (MANET) with many proposals submitted to the IETF for consideration. Delay or Disruption Tolerant Networking (DTN) is a relatively new field for routing, concerned with networks that experience long transmission delay or periods of disruption. Military forces around the globe have applied one or the other networking paradigms with varying degrees of success to their own problems of mobility at the lower tactical level (Brigade and below). The fundamental reason for this limited success is that many of the desired tactical scenarios at this level require a network that is not exclusively ad hoc or exclusively disrupted, but rather a network that dynamically adapts to a variety of mobility situations ranging from relatively stable, almost enterprise like, to completely disrupted. Synchronous MANET protocols have limited disruption tolerance at layer 3, and DTN routing protocols, which are further up the network stack, implement hop by hop asynchronous protocols that are unable to take advantage of prolonged network stability. The primary contribution of this thesis is NOMAD, a new hybrid routing protocol for military mobile ad hoc and disrupted networks. NOMAD is unique in that it operates as a proactive synchronous link state MANET protocol when the network is connected, but is able to seamlessly transition into asynchronous DTN mode when required. The results outlined in this thesis indicate that the hybrid NOMAD protocol provides a substantial improvement over standard synchronous MANET protocols. This thesis also makes a significant contribution with respect to synthetic mobility model generation. Mobility models are essential for the correct evaluation of any routing protocol. A mobility modelling tool called SWarMM (Synthetic Warfare Mobility Modelling) was also developed as part of this thesis. SWarMM provides an agent based simulation tool for generating credible synthetic mobility models for use with the discrete network simulation tools, such as OPNET and NS2.

Cette thèse porte sur l’exploitation d’une population nomade afin de colporter des informations dites ‘atomiques’ de rencontres entre entités fixes ou mobiles pour des applications diverses basées sur la localisation et la traçabilité. Répondre à ces besoins en milieux confinés s’avère un défi, que ce soit dans un contexte industriel, médical ou social. Parmi les solutions existantes, le système GPS (Global Positioning System) offre des résultats encourageants lorsqu’il s’agit de connaître une position en extérieur. Cette solution n’est plus utilisable en intérieur tel que dans une mine par exemple, en raison des caractéristiques du signal fortement perturbées dans ce type de milieu. Le fil directeur de ce travail consiste à proposer un modèle utilisant la technologie des réseaux de capteurs sans fil afin de modéliser et de trouver des solutions à des problématiques allant de la localisation en milieux confinés au suivi et à la reconstitution de trajectoires d’entités mobiles. Les solutions proposées dans cette thématique doivent être tolérantes aux délais. Après avoir présenté les constituants de notre modèle générique de colportage, une instanciation de ce modèle a permis de nous pencher sur un cas de localisation et de traçabilité dans un espace confiné. Les résultats de simulation et d’expérimentation ont montré l’impact de notre mécanisme de colportage sur la qualité de l’information recensée concernant les déplacements et les rencontres des entités mobiles. Outre la proposition d’un modèle générique, la contribution de cette thèse comporte des mécanismes de colportage (et de filtrage) de l’information par des entités mobiles contraintes en taille mémoire. / This thesis focuses on the exploitation of a nomadic population to pass around information called contact event which represents the recording of a meeting between fixed or mobile entities for localization and tracking applications. Addressing these needs in confined environments is a challenge, whether in an industrial, medical or social context. Among the existing solutions, the GPS (Global Positioning System) provides encouraging results for outdoor localization. However, this solution is not operational in confined environments such as mines, due to the signal characteristics highly disturbed in this type of environment. The guiding principle of this work is to propose a model using the technology of wireless sensor networks for modeling and finding solutions for localization and tracking mobile nodes in confined areas. The new approach proposed in this thesis is inspired by the principle of ‘store-carry-forward’ of Delay-Tolerant-Network (DTN). After presenting the components of our generic model, an instantiation of this model has allowed us to address a localization and tracking case in a confined area. The simulation and experimentation results show the impact of our filtering mechanism on the quality of the knowledge collected concerning the movements and meetings of the mobile nodes.

Vehicular ad hoc network (VANET) is an emerging new communication technol-ogy which has attracted a lot of research attention from academic community and indus-try. For many applications in VANETs, information has to be transmitted through mul-tiple hops before it reaches its destination that makes it a subject to various security at-tacks and privacy breaches. Thus, security and privacy issues could limit its adaption by the public community. In this study, we propose and evaluate social based trustworthy data forwarding scheme for VANET. First, by using social network analysis techniques, we provide a framework to strategically deploy Road-Side Units (RSUs) infrastructure in order to im-prove reliability, efficiency, and high packet delivery for VANET. It is based on multiple social centrality assessments of street network which help in placing RSUs at high social intersections. This social placement of RSUs will dramatically improve data dissemina-tion as the opportunity of contacting vehicles increase while costs of RSU deployment can be kept under control. Second, we propose a secure and privacy-preserving message forwarding protocol, which utilizes RSUs to forward messages between vehicles. The protocol takes advantage of high performance capability of RSUs to store and forward messages to their destinations, where these RSUs utilize re-encryption technique to form a mix network to provide adequate privacy for senders and receivers. Then detailed anal-ysis in terms of security, message overhead, delivery ratio, and average delay are per-formed to evaluate the efficiency and effectiveness of our proposed scheme. Lastly, we tackled the security and privacy challenges existing in social-aware data diffusion by proposing an efficient vehicle social evaluation (EVSE) scheme. Our scheme enables each vehicle to show its authentic social evaluation to others while not disclosing its past location information. As a result, it can meet the prerequisites for the success of social aware data diffusion in VANETs. / UOIT

The main principle behind the working of delay tolerant networks (DTN) is the mobility of the nodes along with their contact sequences for exchanging data. Nodes which are a part of the DTN network can behave selfishly due to network reservation policy, especially when constrained to energy or storage space. Several forwarding protocols exist for spreading data but our focus is on the performance of popular data relaying protocols namely epidemic routing and two hop routing protocol in a situation where nodes exhibit various degrees of selfishness. Results of an analytical model show the performance advantage of epidemic routing over two hop routing decreases as the number of selfish nodes and intensity of the selfishness increases either deterministically or probabilistically. We practically asses the vulnerability of the above mentioned protocols using ONE simulator. We find that our result coincides with analytical results with some variations in the graph.

No description available.

Der Beitrag setzt sich mit der halachischen Bedeutung von Dtn. 6,18 im Kontext der heutigen Zeit auseinander. / The article deals with the halakhic importance of Dtn. 6,18 in our times.

Les réseaux tolérants aux délais sont aujourd’hui une alternative prometteuse aux réseaux traditionnels basés sur une infrastructure, encore peu déployée. Il existe plusieurs manières d’évaluer les performances d’un tel réseau : expériences de déploiement grandeur nature, modèles théoriques, simulation, émulation, jeu de traces. Chacune a ses avantages et inconvénients, tant en termes de coûts matériels, de réalisme, de temps nécessaire ou de capacité à gérer des noeuds réels. Cependant, aucune ne répond réellement aux besoins des développeurs d’applications. Dans cette thèse, nous nous focaliserons sur l’émulation. Dans une première partie, nous nous intéresserons aux entrées possibles pour un tel système. Nous proposons tout d’abord un modèle analytique pour prévoir le taux de pertes dans un réseau où les noeuds possèdent une mémoire limitée à un seul paquet. Ensuite, inspirés par les approches de mise à l’échelle de traces de la littérature, nous étudions les hypothèses prises pour l’analyse statistique de traces réelles, et montrons leur influence sur les lois de probabilité obtenues ainsi que les performances réseau observées. Nous étendons ensuite cette étude à la totalité du cycle de vie des traces réelles, en considérant la collecte de données, le filtrage et la mise à l’échelle de celles-ci. Dans une seconde partie, nous proposons une architecture possible d’un émulateur DTN hybride, c’est-à-dire comportant à la fois des noeuds réels sous forme d’intelliphones, et des noeuds virtuels. Le principal avantage ici est de pouvoir évaluer des applications réelles, éventuellement déjà existantes, dans un contexte DTN, et ce de manière aussi transparente que possible. Nous identifions les limites des approches existantes, ce qui nous permet d’établir une liste de spécifications pour notre système. Nous proposons ensuite un système, nommé HINT, permettant de remplir ces spécifications. L’ensemble est ensuite validé, puis appliqué à l’étude de quelques exemples. / Delay Tolerant Networks are currently a promising alternative to infrastructure-based networks, but they have not seen a wide deployment so far. There are several ways to evaluate the performance of such networks: field trials, theoretical models, simulation, emulation or replaying contact datasets. Each one has its advantages and drawbacks in terms of material cost, realism, required time or ability to manage real nodes. However, none of them effectively addresses the needs of application developers. In this thesis, we will focus on emulation. In a first part, we will deal with possible inputs for such a system. We first propose an analytical model to predict the drop ratio in a network where nodes have a one-packet buffer. Then, taking inspiration from trace scaling approaches from the literature, we study the hypotheses and assumptions taken for real traces statistical analyses, showing their impact on the obtained probability distributions and observed network performance metrics. We then extend this study to the whole life cycle of real traces, by considering data collection, filtering and scaling. In a second part, we propose a possible architecture for a hybrid DTN emulator, using both real nodes as smartphones and virtual nodes. The main advantage here is to be able to evaluate real applications, including preexisting ones, in a DTN context, doing so as transparently as possible. We identify the limitations of existing approaches, which helps us build a list of specifications for our system. Then, we propose a system called HINT which matches these specifications. HINT is validated, and applied to the study of some examples.