No description available.

For complex computer networks with many tunable parameters and network performance objectives, the task of selecting the ideal network operating state is difficult. To improve the performance of these kinds of networks, this research proposes the idea of the cognitive network. A cognitive network is a network composed of elements that, through learning and reasoning, dynamically adapt to varying network conditions in order to optimize end-to-end performance. In a cognitive network, decisions are made to meet the requirements of the network as a whole, rather than the individual network components. We examine the cognitive network concept by first providing a definition and then outlining the difference between it and other cognitive and cross-layer technologies. From this definition, we develop a general, three-layer cognitive network framework, based loosely on the framework used for cognitive radio. In this framework, we consider the possibility of a cognitive process consisting of one or more cognitive elements, software agents that operate somewhere between autonomy and cooperation. To understand how to design a cognitive network within this framework we identify three critical design decisions that affect the performance of the cognitive network: the selfishness of the cognitive elements, their degree of ignorance, and the amount of control they have over the network. To evaluate the impact of these decisions, we created a metric called the price of a feature, defined as the ratio of the network performance with a certain design decision to the performance without the feature. To further aid in the design of cognitive networks, we identify classes of cognitive networks that are structurally similar to one another. We examined two of these classes: the potential class and the quasi-concave class. Both classes of networks will converge to Nash Equilibrium under selfish behavior and in the quasi-concave class this equilibrium is both Pareto and globally optimal. Furthermore, we found the quasi-concave class has other desirable properties, reacting well to the absence of certain kinds of information and degrading gracefully under reduced network control. In addition to these analytical, high level contributions, we develop cognitive networks for two open problems in resource management for self-organizing networks, validating and illustrating the cognitive network approach. For the first problem, a cognitive network is shown to increase the lifetime of a wireless multicast route by up to 125\%. For this problem, we show that the price of selfishness and control are more significant than the price of ignorance. For the second problem, a cognitive network minimizes the transmission power and spectral impact of a wireless network topology under static and dynamic conditions. The cognitive network, utilizing a distributed, selfish approach, minimizes the maximum power in the topology and reduces (on average) the channel usage to within 12\% of the minimum channel assignment. For this problem, we investigate the price of ignorance under dynamic networks and the cost of maintaining knowledge in the network. Today's computer networking technology will not be able to solve the complex problems that arise from increasingly bandwidth-intensive applications competing for scarce resources. Cognitive networks have the potential to change this trend by adding intelligence to the network. This work introduces the concept and provides a foundation for future investigation and implementation. / Ph. D.

Dans les réseaux mobiles spontanés (en anglais, Mobile Ad hoc NETworks ou MANETs), la gestion de groupe partitionnable est un service de base permettant la construction d'applications réparties tolérantes au partitionnement. Aucune des spécifications existantes ne satisfait les deux exigences antagonistes suivantes : 1) elle doit être assez forte pour fournir des garanties utiles aux applications réparties dans les systèmes partitionnables ; 2) elle doit être assez faible pour être résoluble. Dans cette thèse, nous proposons une solution à la gestion de groupe partitionnable en environnements réseaux très dynamiques tels que les MANETs. Pour mettre en œuvre notre solution, nous procédons en trois étapes. Tout d'abord, nous proposons un modèle de système réparti dynamique qui caractérise la stabilité dans les MANETs. Ensuite, nous adaptons pour les systèmes partitionnables l'approche Paxos à base de consensus Synod. Cette adaptation résulte en la spécification d'un consensus abandonnable AC construit au-dessus d'un détecteur ultime des α participants d'une partition ♢PPD et d'un registre ultime par partition ♢RPP. ♢PPD garantit la vivacité dans une partition même si la partition n'est pas complètement stable tandis que ♢RPP préserve la sûreté dans la même partition. Enfin, la gestion de groupe partitionnable est résolue en la transformant en une séquence d'instances de AC. Chacun des modules ♢PPD, ♢RPP, AC et gestion de groupe partitionnable est implanté et prouvé. Par ailleurs, nous analysons les performances de ♢PPD par simulation / In Mobile Ad hoc NETworks or MANETs, partitionable group membership is a basic service for building partition-tolerant applications. None of the existing specifications satisfy the two following antagonistic requirements: 1) it must be strong enough to simplify the design of partition-tolerant distributed applications in partitionable systems; 2) it must be weak enough to be implantable. In this thesis, we propose a solution to partitionable group membership in very dynamic network environment such as MANETs. To this means, we proceed in three steps. First, we develop a dynamic distributed system model that characterises stability in MANETs. Then, we propose a solution to the problem of partitionable group membership by adapting Paxos for such systems. This adatation results in a specification of abortable consensus AC which is composed of an eventual α partition-participants detector ♢PPD and an eventual register per partition ♢RPP. ♢PPD guarantees liveness in a partition even if the partition is not completely stable whereas ♢RPP ensures safety in the same partition. Finally, partitionable group membership is solved by transforming it into a sequence of abortable consensus instances AC. Each of the modules ♢PPD, ♢RPP, AC, and partitionable group membership is implanted and proved. Next, we analyse the performances of ♢PPD by simulation

L’intérêt porté aux algorithmes probabilistes est, entre autres,dû à leur simplicité. Cependant, leur analyse peut devenir très complexeet ce particulièrement dans le domaine du distribué. Nous mettons en évidencedes algorithmes, optimaux en terme de complexité en bits résolvantles problèmes du MIS et du couplage maximal dans les anneaux, qui suiventle même schéma. Nous élaborons une méthode qui unifie les résultatsde bornes inférieures pour la complexité en bits pour les problèmes duMIS, du couplage maximal et de la coloration. La complexité de ces analysespouvant facilement mener à l’erreur et l’existence de nombreux modèlesdépendant d’hypothèses implicites nous ont motivés à modéliserde façon formelle les algorithmes distribués probabilistes correspondant ànotre modèle (par passage de messages, anonyme et synchrone), en vuede prouver formellement des propriétés relatives à leur analyse. Pour cela,nous développons une bibliothèque, RDA, basée sur l’assistant de preuveCoq. / Probabilistic algorithms are simple to formulate. However, theiranalysis can become very complex, especially in the field of distributedcomputing. We present algorithms - optimal in terms of bit complexityand solving the problems of MIS and maximal matching in rings - that followthe same scheme.We develop a method that unifies the bit complexitylower bound results to solve MIS, maximal matching and coloration problems.The complexity of these analyses, which can easily lead to errors,together with the existence of many models depending on implicit assumptionsmotivated us to formally model the probabilistic distributed algorithmscorresponding to our model (message passing, anonymous andsynchronous). Our aim is to formally prove the properties related to theiranalysis. For this purpose, we develop a library, called RDA, based on theCoq proof assistant.

Wireless sensor networks (WSNs) have recently achieved a great deal of attention due to its numerous attractive applications in many different fields. Sensors and WSNs possesses a number of special characteristics that make them very promising in many applications, but also put on them lots of constraints that make issues in sensor network particularly difficult. These issues may include topology control, routing, coverage, security, and data management. In this thesis, we focus our attention on the coverage problem. Firstly, we define the Sensor Energy-efficient Scheduling for k-coverage (SESK) problem. We then solve it by proposing a novel, completely localized and distributed scheduling approach, naming Distributed Energy-efficient Scheduling for k-coverage (DESK) such that the energy consumption among all the sensors is balanced, and the network lifetime is maximized while still satisfying the k-coverage requirement. Finally, in related work section we conduct an extensive survey of the existing work in literature that focuses on with the coverage problem.

Wireless Sensor Networks (WSNs) comprise a large number of sensor nodes, which are spread out within a region and communicate using wireless links. In some WSN applications, recognizing boundary nodes is important for topology discovery, geographic routing and tracking. In this thesis, we study the problem of identifying the boundary nodes of a WSN. In a WSN, close-by nodes can communicate with their neighbors and have the ability to estimate distances to nearby nodes, but not necessarily the true distances. Our objective is to find the boundary nodes by using the connectivity relation and neighbor distance information without any knowledge of node locations. Moreover, our main aim is to design a distributed algorithm that works even when the average degree is low. We propose a heuristic algorithm to find the boundary nodes which are connected in a boundary cycle of a location-free, low density, randomly deployed WSN. We develop the key ideas of our boundary detection algorithm in the centralized scenario and extend these ideas to the distributed scenario. Then, we show by simulation experiments that the distributed implementation outperforms the centralized one. The centralized implementation relies on the connectivity of the network to the base station. Therefore, for low density disconnected networks, the algorithm cannot find boundaries in partitions of the network that cannot establish connection to the base station. This condition leads to a low quality of boundary discovery. In contrast, the distributed implementation is more realistic for real WSNs, especially for relatively sparse networks when all local information cannot be collected very well due to sparse connectivity. In low-degree disconnected networks, the simulation results show that the distributed implementation has a higher quality of boundaries compared to the centralized implementation.

Ad hoc networks are emerging as a cost-effective, yet, powerful tool for communication. These systems, where networks can emerge and converge on-the-fly, are guided by the forward-looking goals of providing ubiquitous connectivity and constant access to information. Due to power and bandwidth constraints, the vulnerability of the wireless medium, and the multi-hop nature of ad hoc networks, these networks are becoming increasingly complex dynamic systems. Besides, modern radios are empowered to be reconfigurable, which harbors the temptation to exploit the system. To understand the implications of these issues, some of which pose significant challenges to efficient network design, we study topology control using game theory. We develop a game-theoretic framework of topology control that broadly captures the radio parameters, one or more of which can be tuned under the purview of topology control. In this dissertation, we consider two parameters, viz. transmit power and channel, and study the impact of controlling these on the emergent topologies. We first examine the impact of node selfishness on the network connectivity and energy efficiency under two levels of selfishness: (a) nodes cooperate and forward packets for one another, but selfishly minimize transmit power levels and; (b) nodes selectively forward packets and selfishly control transmit powers. In the former case, we characterize all the Nash Equilibria of the game and evaluate the energy efficiency of the induced topologies. We develop a better-response-based dynamic that guarantees convergence to the minimal maximum power topology. We extend our analysis to dynamic networks where nodes have limited knowledge about network connectivity, and examine the tradeoff between network performance and the cost of obtaining knowledge. Due to the high cost of maintaining knowledge in networks that are dynamic, mobility actually helps in information-constrained networks. In the latter case, nodes selfishly adapt their transmit powers to minimize their energy consumption, taking into account partial packet forwarding in the network. This work quantifies the energy efficiency gains obtained by cooperation and corroborates the need for incentivizing nodes to forward packets in decentralized, energy-limited networks. We then examine the impact of selfish behavior on spectral efficiency and interference minimization in multi-channel systems. We develop a distributed channel assignment algorithm to minimize the spectral footprint of a network while establishing an interference-free connected network. In spite of selfish channel selections, the network spectrum utilization is shown to be within 12% of the minimum on average. We then extend the analysis to dynamic networks where nodes have incomplete network state knowledge, and quantify the price of ignorance. Under the limitations on the number of available channels and radio interfaces, we analyze the channel assignment game with respect to interference minimization and network connectivity goals. By quantifying the interference in multi-channel networks, we illuminate the interference reduction that can be achieved by utilizing orthogonal channels and by distributing interference over multiple channels. In spite of the non-cooperative behavior of nodes, we observe that the selfish channel selection algorithm achieves load balancing. Distributing the network control to autonomous agents leaves open the possibility that nodes can act selfishly and the overall system is compromised. We advance the need for considering selfish behavior from the outset, during protocol design. To overcome the effects of selfishness, we show that the performance of a non-cooperative network can be enhanced by appropriately incentivizing selfish nodes. / Ph. D.

Nous proposons tout d'abord une étude de plusieurs problèmes de l'algorithmique distribuée. Nous fournissons un modèle formel appliqué aux réseaux de diffusion anonymes. Dans ce modèle, nous caractérisons les graphes dans lesquels il est possible de résoudre l'énumération et l'élection. Cette caractérisation se base sur la notion d'homomorphisme de graphes. Nous proposons deux algorithmes dont la complexité est polynomiale et qui améliorent les complexités exponentielles connues jusqu'à présent. Dans un second temps, nous étudions le problème du calcul de l'état global et nous introduisons la notion de weak snapshot. Nous montrons qu'il existe des solutions pour ce problème dans les réseaux anonymes. Nous présentons plusieurs résultats concernant le calcul de l'état global en liaison avec des applications telles que le calcul de points de reprise, la détection de la terminaison ou encore le calcul d'une cartographie du réseau. Dans un cadre plus pratique, nous présentons la conception, le développement et l'implémentation des algorithmes proposés pour le calcul de l'état global au sein du logiciel de simulation et de visualisation ViSiDiA. / In this thesis, we first present a study of several problems in the field of distributed algorithms. We provide a formal model that relies on anonymous networks. In this model, we characterize graphs in which it is possible to solve enumeration and leader election problems. This characterization is based on graph homomorphism. We introduce two algorithms with polynomial complexities that improve existing works with exponential complexities. On the other hand, we study the snapshot problem and we introduce the notion of weak snapshot. We show that there exist solutions for this problem in the context of anonymous networks. We present several results about distributed snapshots that deal with checkpoint and rollback recovery, termination detection or the cartography computation of a network. In a practical aspect, we present the conception, the development process and the implementation of these distributed snapshot algorithms within the simulation and visualization software ViSiDiA.

Edsger W. Dijkstra prägte im Jahr 1974 den Begriff Selbststabilisierung (self-stabilization) in der Informatik. Ein System ist selbststabilisierend, wenn es von jedem denkbaren Zustand aus nach einer endlichen Anzahl von Aktionen ein stabiles Verhalten erreicht. Im Mittelpunkt dieser Arbeit steht der Entwurf selbststabilisierender Algorithmen. Wir stellen eine Petrinetz-basierte Methode zum Entwurf selbststabilisierender Algorithmen vor. Wir validieren unsere Methode an mehreren Fallstudien: Ausgehend von algorithmischen Ideen existierender Algorithmen beschreiben wir jeweils die die schrittweise Entwicklung eines neuen Algorithmus. Dazu gehört ein neuer randomisierter selbststabilisierender Algorithmus zur Leader Election in einem Ring von Prozessoren. Dieser Algorithmus ist abgeleitet aus einem publizierten Algorithmus, von dem wir hier erstmals zeigen, daß er fehlerhaft arbeitet. Wir weisen die Speicherminimalität unseres Algorithmus nach. Ein weiteres Ergebnis ist der erste Algorithmus, der ohne Time-Out-Aktionen selbststabilisierenden Tokenaustausch in asynchronen Systemen realisiert. Petrinetze bilden einen einheitlichen formalen Rahmen für die Modellierung und Verifikation dieser Algorithmen. / In 1974, Edsger W. Dijkstra suggested the notion of self-stabilization. A system is self-stabilizing if regardless of the initial state it eventually reaches a stable behaviour. This thesis focuses on the design of self-stabilizing algorithms. We introduce a new Petri net based method for the design of self-stabilizing algorithms. We validate our method on several case studies. In each of the case studies, our stepwise design starts from an algorithmic idea and leads to a new self-stabilizing algorithm. One of these algorithms is a new randomized self-stabilizing algorithm for leader election in a ring of processors. This algorithm is derived from a published algorithm which we show to be incorrect. We prove that our algorithm is space-minimal. A further result is the first algorithm for token-passing in a asynchronous environment which works without time-out actions. Petri nets form a unique framework for modelling and verification of these algorithms.

In this dissertation, we focus on two related parts of a 3D face recognition system with wireless transportation. In the first part, the core components of the system, namely, the feature extraction and classification component, are introduced. In the feature extraction component, range images are taken as inputs and processed in order to extract features. The classification component uses the extracted features as inputs and makes classification decisions based on trained classifiers. In the second part, we consider the wireless transportation problem of range images, which are captured by scattered sensor nodes from target objects and are forwarded to the core components (i.e., feature extraction and classification components) of the face recognition system. Contrary to the conventional definition of being a transducer, a sensor node can be a person, a vehicle, etc. The wireless transportation component not only brings flexibility to the system but also makes the “proactive” face recognition possible. For the feature extraction component, we first introduce the 3D Morphable Model. Then a 3D feature extraction algorithm based on the 3D Morphable Model is presented. The algorithm is insensitive to facial expression. Experimental results show that it can accurately extract features. Following that, we discuss the generic face warping algorithm that can quickly extract features with high accuracy. The proposed algorithm is robust to holes, facial expressions and hair. Furthermore, our experimental results show that the generated features can highly differentiate facial images. For the classification component, a classifier based on Mahalanobis distance is introduced. Based on the classifier, recognition performances of the extracted features are given. The classification results demonstrate the advantage of the features from the generic face warping algorithm. For the wireless transportation of the captured images, we consider the location-based wireless sensor networks (WSN). In order to achieve efficient routing perfor¬mance, a set of distributed stateless routing protocols (PAGER) are proposed for wireless sensor networks. The loop-free and delivery-guaranty properties of the static version (PAGER-S) are proved. Then the performance of PAGER protocols are compared with other well-known routing schemes using network simulator 2 (NS2). Simulation results demonstrate the advantages of PAGER.