Computação em grade tem se estabelecido como um importante paradigma de computação, por permitir lidar com grandes quantidades de cálculos e dados e a colaboração de participantes geograficamente distribuídos. Esses sistemas devem ser organizados de forma completamente distribuídas, com cada participante mantendo informações sobre outros participantes, e as informações necessárias ao funcionamento do sistema circulando pela rede de overlay resultante. Quando novas propostas de algoritmos, protocolos ou infraestruturas para a grade são apresentadas, sua avaliação efetiva implica considerar sua operação com uma grande quantidade de participantes, o que invariavelmente significa que simulações devem ser realizadas. Este trabalho apresenta um sub-sistema de simulação de redes de overlay integrado à plataforma de simulação de computação de grade GridSim, de forma a facilitar o estudo desse tipo de estruturas e o desenvolvimento de novas propostas de protocolos e algoritmos para seu uso em grades de computadores. A metodologia adotada resultou no desenvolvimento de um Java package no GridSim com classes e interfaces que representam os conceitos básicos de redes de overlay e da interface dos clientes com essas redes. A partir dele foi possível desenvolver protocolos para redes estruturadas e não estruturadas no simulador e simulá-los utilizando cenários de grade de dados. Com os resultados obtidos foi possível observar que, os protocolos implementados no simulador estão de acordo com o que é encontrado na literatura. / Grid Computing has been established as an important computing paradigm, since it allows dealing with a large quantity of computations and data and the collaboration of geographically distributed participants. Those systems must be organized in a completely distributed way, with each participant knowing about some other participants, and the needed information to the functioning system circulating through the resulting overlay network. When new algorithm proposals, protocols or infrastructures to the grid are presented, its evaluation implies to consider its operation with a large number of participants, which invariably means that simulations must be done. This work presents a subsystem of overlay network simulation integrated to the GridSim simulation platform, in order to facilitate the study of that type of structures and the development of new protocols and algorithms for use in grid computers. The adopted methodology led up to the development of a Java package with classes and interfaces that represent the basic concepts of overlay networks and of the clients interface with those networks. Using this package, it was possible for develop protocols to structured and non-structured networks in the simulator and simulate them using data grid scenarios. With the obtained results it was possible to observe that the implemented protocols in the simulator agree with what is found in the literature.

Content delivery through service overlay networks has gained popularity due to the overlays’ abilities to provide effective and reliable services. Inefficiencies of one-to-one matching of user requirements to a single service have given rise to service composition. Customized media delivery can be achieved through dynamic compositions of Service Specific Overlay Networks (SSONs). However, the presence of SSONs in dynamic environments raises the possibility of unexpected failures and quality degradations. Thus constructing, managing, and repairing corrupted service paths are challenging dilemmas. This thesis investigates the problem of autonomous SSON construction and management and identifies the drawbacks of current approaches. A novel multi-layered, autonomous, self-adaptive framework for constructing SSONs is presented. The framework includes a Hybrid Service Overlay Network layer (H-SON). The H-SON is a dynamic hybrid overlay dedicated to service composition for multimedia delivery in dynamic networks. Node placement in the overlay depends on the node’s stability, types and quality of provided services. Changes in stability and QoS of service nodes are reflected by dynamic re-organizations of the overlay. The H-SON permits fast and efficient searches for component services that meet client functional and quality expectations. Self-managed overlay nodes coordinate their behaviors to formulate a service composition path that meets the client’s requirements. Two approaches are presented in this work. The first illustrates how SSONs are established through dynamically adaptable MS-designed plans. The imprecise nature of nonfunctional service characteristics, such as QoS, is modeled using a fuzzy logic system. Moreover, semantic similarity evaluations enable us to include, in compositions, those services whose operations match, semantically, the requirements of the composition plan. Plan-based composition solutions restrict service discovery to defined abstract models. Our second composition approach introduces a semantic similarity and nearness SSON composition method. The objective is to free service nodes from the adherence to restrictive composition plans. The presented work illustrates a service composition solution that semantically advances service composition paths towards meeting users’ needs with each service hop while simultaneously guaranteeing user-acceptable QoS levels. Simulation results showcase the effectiveness of the presented work. Gathered results validate the success of our service composition methods while meeting user requirements.

In today's Internet, building overlay structures to provide a service is becoming more and more common. This approach allows for the utilization of client resources, thus being more scalable than a client-server model in this respect. However, in these architectures the quality of the provided service depends on the clients and is therefore more complex to manage. Resource utilization, both at the clients themselves and in the underlying network, determine the efficiency of the overlay application. Here, a trade-off exists between the resource providers and the end users that can be tuned via overlay mechanisms. Thus, resource management and traffic management is always quality-of-service management as well. In this monograph, the three currently significant and most widely used overlay types in the Internet are considered. These overlays are implemented in popular applications which only recently have gained importance. Thus, these overlay networks still face real-world technical challenges which are of high practical relevance. We identify the specific issues for each of the considered overlays, and show how their optimization affects the trade-offs between resource efficiency and service quality. Thus, we supply new insights and system knowledge that is not provided by previous work. / Im heutigen Internet werden immer häufiger Overlay-Strukturen aufgebaut, um eine Dienstleistung zu erbringen. Dieser Ansatz ermöglicht die Nutzung von Client-Ressourcen, so dass er in dieser Hinsicht besser skaliert als das Client-Server-Modell. Die Qualität des zur Verfügung gestellten Dienstes hängt nun aber von den Clients ab und ist daher komplizierter zu steuern. Die Ressourcennutzung, sowohl auf den Clients selbst als auch in dem zugrunde liegenden Netzwerk, bestimmt die Effizienz der Overlay-Anwendung. Hier existiert ein Trade-off zwischen Ressourcen-Anbietern und Endkunden, der über Overlay-Mechanismen geregelt werden kann. Daher ist Ressourcenmanagement und Traffic-Management gleichzeitig immer auch Quality-of-Service-Management. In dieser Arbeit werden die drei derzeit am weitesten im Internet verbreiteten und signifikanten Overlay-Typen berücksichtigt. Diese Overlays sind in populären Anwendungen, die erst vor kurzem an Bedeutung gewonnen haben, implementiert. Daher sind diese Overlay-Netze nach wie vor realen technischen Herausforderungen ausgesetzt, die von hoher praktischer Relevanz sind. Die spezifischen Herausforderungen für jedes der betrachteten Overlays werden identifiziert und es wird gezeigt, wie deren Optimierung den Trade-off zwischen Ressourceneffizienz und Service-Qualität beeinflusst. So werden neue Einsichten und Erkenntnisse über diese Systeme gewonnen, die in früheren Arbeiten nicht existieren.

The research presented in this thesis covers two critical problems regarding lined pipes: dynamic impact and welding. A lined pipe consists of an inner layer (the liner) made of corrosion resistant alloy (CRA), e.g. AISI304 stainless steel, and an outer layer made of low carbon steel, e.g. carbon-manganese steel, C-Mn. To manufacture the lined pipe, a special heat treatment, known as tight fit pipe (TFP), based on cooling the liner to -200°C, heating the backing pipe to +500°C and inserting the liner inside the outer pipe, was used in this work. Both welding and impact with external objects are responsible for accumulating high levels of plastic strains and residual stresses which could lead to failure in the pipe sometime after the impact or the welding. The special welding process used in lined pipes typically consists of the overlay welding (inner welding) of the liner with the C-Mn steel pipe for each segment and the girth welding (outer welding) of the two segments. To simulate this welding process using the ABAQUS code, nonlinear heat-transfer and mechanical finite-element (FE) analyses have been conducted. A distributed power density of the moving welding torch and a non-linear heat transfer coefficient accounting for both radiation and convection have been used in the analysis and implemented in ABAQUS user-subroutines. The modelling procedure has been validated first against previously published experimental results for stainless steel and carbon steel pipe welding separately. The model has been then used to determine the isotherms induced by the one-pass weld overlay and the one-pass girth welding and to clarify their influence on the transient temperature field and residual stress in the lined pipe. Furthermore, the influence of the cooling time between weld overlay and girth welding and of the welding speed have been examined thermally and mechanically as they are key factors that can affect the quality of lined pipe welding. The same FE numerical procedure to analyse line pipe welding is then applied to simulate six cases experimentally tested in the lab within this project. Furthermore, two cases have been analysed first, namely a reference case, in which the effect of the TFP pre-heat treatment is neglected, and a second one where the pre-heat treatment has been taken into consideration. During welding, the FE thermal history and mechanical strain results for both cases correlate well with the experimental ones in the region with the highest residual stresses, because the effect of initial residual stresses is cancelled in the regions subject to very high temperatures. After welding, the numerical and experimental results have proved that the initial residual stresses due to the TFP pre-heat treatment are reasonably important in the liner whereas they are practically negligible in the C-Mn pipe. The same reference case is then compared numerically and experimentally with further five parametric cases to study the effect of welding properties (weld overlay and girth welding materials), geometric parameters (using weld overlay and liner) and welding process parameters (heat input). The numerical temperature fields and residual stresses are in good agreement with their experimental counterparts for all cases. The dynamic impact problem is a crucial one for lined pipes because of the reduction in the thickness of the outer pipe ensured by the internal protection from corrosion given by a thinner liner. In this case, the lined pipe is more affected by potential impact with external objects (so-called 'third party interference' in the Oil and Gas industry). In general, a dent produced by a freely dropped weight is responsible to a large extent of catastrophic failure in pipelines. Therefore, in this work, 3D FE models have been developed to simulate the mechanism of vertical free drop of a weight from different heights resulting in damage in the pipe. Models have been executed using a three-dimensional non-linear explicit-dynamics FE code, ABAQUS/EXPLICIT. In order to precisely simulate the response of the pipe to subsequent impacts and spring back, an elastic-plastic constitutive law is adopted using the isotropic Hooke's law and a Von Mises yield criterion, with work hardening based on an isotropic hardening rule associated with the equivalent plastic strain rate. Strain-rate dependent properties are specified for both materials, C-Mn and AISI304, to take into account the change in velocities during impact. The numerical strain results are reasonably consistent with the experimental ones recorded by four strain gauge rosettes positioned symmetrically around the dent centre. Numerical and experimental results are comprehensively analysed and discussed also in terms of practical implications in the industry.

Today's distributed systems exist on a scale that was unimaginable only a few decades ago. Distributed systems now can consist of thousands or even millions of computers spread across the entire world. These large systems are often organized into overlay networks - networks composed of virtual links, with each virtual link realized by one or more physical links. Self-stabilizing overlay networks promise that, starting from any weakly-connected configuration, the correct network topology is always built. This area of research is young, and prior examples of self-stabilizing overlay networks have either been for simple topologies, or involved complex algorithms that were difficult to verify and extend. We address these limitations in this thesis. First, we present the Transitive Closure Framework, a generic framework to transform any locally-checkable overlay network into a self-stabilizing network. This simple framework has a running time which is at most a logarithmic number of rounds more than optimal, and in fact is optimal for a particular class of overlay networks. We also prove the only known non-trivial lower bound on the convergence time of any self-stabilizing overlay network. To allow fast and efficient repairs for local faults, we extend the Transitive Closure Framework to the Local Repair Framework. We demonstrate this framework by implementing an efficient algorithm for node joins in the Skip+ graph. Next, we present the Avatar network, which is a generic locally checkable overlay network capable of simulating many other overlay networks. We design a self-stabilizing algorithm for a binary search tree embedded onto the Avatar network, and prove this algorithm requires only a polylogarithmic number of rounds to converge and limits degree increases to within a polylogarithmic factor of optimal. This algorithm is the first to achieve such efficiency, and its modular design makes it easy to extend. Finally, we introduce a technique called network scaffolding, which builds other overlay network topologies using the Avatar network.

Avec la diminution constante de la taille des transistors dans la microélectronique, les outils de caractérisation doivent être de plus en plus précis et doivent fournir un débit de plus en plus élevé. La fabrication de semi-conducteurs étant un processus couche par couche, le positionnement précis de la pile est crucial. Le mauvais alignement de la pile est appelé overlay, et nous proposons ici un nouvel instrument et une nouvelle méthode pour caractériser avec précision l'overlay en mesurant une cible unique construite dans les lignes de découpe. La méthode utilise les propriétés fondamentales de symétrie de la matrice de Mueller mesurée dans le plan focal arrière d'un objectif de microscope à grande ouverture numérique et permet une caractérisation de l'overlay avec une incertitude de mesure totale de 2nm. Après une brève introduction à la polarisation et la matrice de Mueller, nous décrivons la nouvelle conception de l'instrument et son étalonnage complet. Le corps principal de ce manuscrit est dédié à la caractérisation de l'overlay, mais les applications de cet instrument sont très diverses aussi détaillerons nous comment notre instrument peut apporter des pistes pour la caractérisation et la compréhension de l'auto-organisation de l'exosquelette des scarabées. Ces coléoptères présentent un très fort dichroïsme circulaire et de nombreux groupes de recherche dans le monde entier essaient d'imiter leur exosquelette. Nous concluons ce manuscrit par un bref aperçu des principales perspectives pour notre instrument.

Les réseaux de recouvrement ont été popularisés avec l'expension des réseaux pair-à-pair (peer-to-peer networks) dans les années 2000. Il existe de nombreux types de réseaux de recouvrement, certains sont extensibles d'autres non, certains ont pour rôle d'assurer une topologie résiliente, certain offre un service de routage fiable, etc. Mais aucun des réseaux de recouvrement n'est adaptables à différents types de réseaux recouvert et ne proposer un large éventail de services. Tout réseau de recouvrement existant est uniquement dédié à une tâche particulière. On peut imaginer un réseau qui pourrait être déployé à cheval sur de nombreux types différents de réseau et ne se consacrant pas seulement à une tâche particulière. Cette thèse a pour objectif de définir les bases et de développer un tel réseau de recouvrement. Dans ces travaux de recherches, nous proposons un nouveau réseau de recouvrement appelé ROSA. ROSA est conçu de façon à pouvoir facilement être adapté aux différents réseaux physiques et d'être en mesure de fournir un large éventail de services différents. Les nœuds de ROSA sont organisés en cluster appelé grumeaux (lump) et ROSA peut être considéré comme un enchevêtrement de grumeaux. Les nœuds organisent leurs ensembles de voisin en fonction des densités associées à ces grumeaux. ROSA est extensible car le nombre maximal de voisins qu'un noeud peut avoir est borné, et cette limite ne dépend pas du nombre total de nœuds participant au réseau. ROSA est adaptable car la définition de la densité des grumeaux peuvent être modifies et adaptés aux propriétés du réseau recouvert. La densité définit le comportement de ROSA

The constant growth in network communications technologies and the emergence of Service Specific Overlay Networks (SSONs), coupled with the rapid development of multimedia applications make the management of such technologies a major challenge. This thesis investigates the SSONs management problem and proposes an autonomic architecture, a self-organizing and self-adapting algorithm, and a utility function for monitoring the Quality of Experience (QoE) of IPTV streams in SSONs. First, we examine the different issues stemming from the autonomic management of SSONs and identify the limitations of existing approaches. We then propose an architecture to ease the management of SSONs by incorporating autonomic computing principles to make SSONs acquire self-management capabilities. The proposed architecture introduces autonomic control loops that continuously monitor network components and analyze the gathered data. An Autonomic System (AS) is comprised of one or more Autonomic Managers (AM) which take control of managing other elements in the network. The proposed architecture highlights the different components of an AM and identifies its purpose. The distributed nature of the proposed architecture avoids limitations of centralized management solutions. We then propose a scheme to allow AMs to emerge among the set of nodes in the network as the most powerful ones in terms of different factors, including processing capabilities and stability. Using a self-organizing and self-adapting distributed protocol, each node in the overlay selects an appropriate AM to report to so that sensed data is delivered error-free, and in a timely manner, while the load is distributed over the AMs. Finally, we propose a utility function to monitor the quality of IPTV streams by predicting QoE based on statistical Quality of Service (QoS) information. The proposed function is simple and does not require high processing power. It allows the QoE of IPTV users to be monitored in real-time by the AMs, so that quality degradations are accurately identified and adaptation mechanisms are triggered at the right moment to correct issues causing degradations. Theoretical analysis and simulations studies are presented to demonstrate the performance of the proposed schemes.

While the demand for video streaming services has risen rapidly in recent years, supporting video streaming service to a large number of receivers still remains a challenging task. Issues of video streaming in the Internet, such as scalability, and reliability are still under extensive research. Recently proposed network contexts such as overlay networks, and mobile ad hoc networks pose even tougher challenges. This thesis focuses on supporting scalable video streaming applications under various network environments. More specifically, this thesis investigates the following problems: i) Server selection in replicated batching video on demand (VoD) systems: we find out that, to optimize the user perceived latency, it is vital to consider the server state information and channel allocation schemes when making server selection decisions. We develop and evaluate a set of server selection algorithms that use increasingly more information. ii) Scalable live video streaming with time shifting and video patching: we consider the problem of how to enable continuous live video streaming to a large group of clients in cooperative but unreliable overlay networks. We design a server-based architecture which uses a combined technique of time-shifting video server and P2P video patching. iii) A Cooperative patching architecture in overlay networks: We design a cooperative patching architecture which shifts video patching responsibility completely to the client side. An end-host retrieves lost data from other end-hosts within the same multicast group. iv) V3: a vehicle to vehicle video streaming architecture: We propose V3, an architecture to provide live video streaming service to driving vehicles through vehicle-to-vehicle (V2V) networks. V3 incorporates a novel signaling mechanism to continuously trigger video sources to send video data back to the receiver. It also adopts a store-carry-and-forward approach to transmit video data in a partitioned network environment. We also develop a multicasting framework that enables live video streaming applications from multiple sources to multiple receivers in V2V networks. A message integration scheme is used to suppress the signaling overhead, and a two-level tree-based routing approach is adopted to forward the video data.

Configuration capabilities are important for modern advanced network services. Network conditions and user populations have been significantly diversified after decades of evolution of the Internet. Configuration capabilities allow network services to be adapted to spatial, temporal, and managerial variations in application requirements and service operation conditions. Network service providers need to decide on the best configuration. Ideally, a network service should have all of its components optimally configured to most effectively deliver the functionality for which it was designed. The optimal configuration, however, is always a compromise between different metrics. To decide on an optimal configuration, the prominent performance and cost metrics must be identified, modeled, and quantified. Optimization objective functions and constraints that combine these metrics should be formulated and optimization techniques should be developed. More important, in the scenarios where the application requirements and system conditions change over time, the service configuration needs to be dynamically adjusted and strategies that guide the reconfiguration decisions need to be developed. Because the actual process of configuring a network service incurs configuration costs, an optimal reconfiguration strategy should be one that achieves a tradeoff between the (re)configuration costs and static optimization objectives. Furthermore, such tradeoffs must be based on the consideration of long-term benefits instead of short-term interest. This thesis focuses on understanding the strategies for dynamic (re)configuration of advanced network services positioned above the Transport Layer. Specifically, this thesis investigates the configuration and more important dynamic reconfiguration strategies for two types of advanced network services: Service Overlay Networks, and Content Resiliency Service Networks. Unlike those network services whose configuration involves mainly arrangement of hard-wired components, these network services have the ability to change service configuration in small time scales. This makes the modeling of application requirements and system condition dynamics not only possible but also meaningful and potentially useful. Our goal is to develop modeling and optimization techniques for network service configuration and dynamic reconfiguration policies. We also seek to understand how effective techniques can improve the performance or reduce the cost of these advanced network services, thus demonstrating the advantage of allowing configurability in these advanced network services.