Submitted by Setor de Tratamento da Informa??o - BC/PUCRS (tede2@pucrs.br) on 2016-07-18T20:23:46Z No. of bitstreams: 1 DIS_EVERTON_DE_MATOS_COMPLETO.pdf: 1251543 bytes, checksum: f9774bc8a409974553769c0c5777f444 (MD5) / Made available in DSpace on 2016-07-18T20:23:46Z (GMT). No. of bitstreams: 1 DIS_EVERTON_DE_MATOS_COMPLETO.pdf: 1251543 bytes, checksum: f9774bc8a409974553769c0c5777f444 (MD5) Previous issue date: 2016-03-29 / The computing paradigm called Internet of Things (IoT) will connect billions of devices deployed around the world together in a near future. In the last years, IoT is gaining more attention. This paradigm has become popular by embedding mobile network and processing power into a wide range of physical computing devices used in everyday life of many people. An important element of the IoT is a middleware, which is a system able to abstract the management of physical devices and to provide services based on the information from these devices. The services provided are used by application clients to perform queries and obtain environmental information. In this way, it is already a subject in literature studies that address middleware systems not only interoperability of devices, but also context awareness feature. Context-aware is an important feature of IoT systems. This feature makes easy to discover, understand, and store relevant information related to devices. This information can be used for a refined provision of services based on the environment context and also for decision making. This work aims to present the Context-Aware System (CONASYS), that is a system to provide services of contextualized information about IoT devices in heterogeneous environments. The system is attached to COMPaaS IoT middleware and is able to act accordingly to the environment that it is inserted. We present in details the architecture of CONASYS, the technical issues related to the implementation of the system and perform some tests based in a real-world scenario. We also present some related work. Our objective is to provide a well-defined range of contextualized services that meet the users needs without specific knowledge of the environment, improving users Quality of Experience (QoE). / O paradigma da computa??o Internet das Coisas (IoT) ir? conectar bilh?es de dispositivos ao redor do mundo em um futuro pr?ximo. Nos ?ltimos anos, a IoT vem ganhando mais aten??o. Esse paradigma tornou-se popular por embarcar redes m?veis e pelo seu poder de processamento em uma vasta gama de dispositivos computadorizados utilizados na vida cotidiana de muitas pessoas. Um elemento importante da IoT ? o middleware, que ? um sistema capaz de abstrair a gest?o de dispositivos e prover servi?os baseados nestes dispositivos. Os servi?os providos s?o usados por aplica??es para obter informa??es do ambiente. Desta forma, existem muitas pesquisas relacionadas com o desenvolvimento de middleware que abordam n?o s? interoperabilidade dos dispositivos, mas tamb?m a caracter?stica de ci?ncia de contexto. Ci?ncia de contexto ? uma caracter?stica importante dos sistemas da IoT. Esta caracter?stica facilita o descobrimento, compreens?o e armazenamento de informa??es relevantes relacionadas aos dispositivos. Estas informa??es podem ser usadas para prover servi?os e tomada de decis?o com base no contexto do ambiente. para provimento de servi?os de informa??o contextulizada sobre dispositivos da IoT em ambientes heterog?neos. O sistema ? acoplado ao middleware COMPaaS e ? capaz de agir conforme o ambiente que est? inserido. A arquitetura do CONASYS ? apresentada em detalhes, assim como os testes realizados. Nosso objetivo ? prover servi?os contextualizados que atendam ?s necessidades dos usu?rios que n?o possuem conhecimento espec?fico do ambiente, melhorando assim a Qualidade da Experi?ncia (QoE).

O crescente desenvolvimento de aplicações de alto desempenho é uma realidade presente nos dias atuais. Entretanto, a diversidade de arquiteturas de máquinas, incluindo monoprocessadores e multiprocessadores, clusters com ou sem máquina front-end, variedade de sistemas operacionais e implementações da biblioteca MPI tem aumentado cada dia mais. Tendo em vista este cenário, bibliotecas que proporcionem a integração de diversas implementações MPI, sistemas operacionais e arquiteturas de máquinas são necessárias. Esta tese apresenta o HyMPI, uma implementação da biblioteca MPI voltada para integração, em um mesmo ambiente distribuído de alto desempenho, nós com diferentes arquiteturas, clusters com ou sem máquina front-end, sistemas operacionais e implementações MPI. HyMPI oferece um conjunto de primitivas compatíveis com a especificação MPI, incluindo comunicação ponto a ponto, operações coletivas, inicio e termino, além de outras primitivas utilitárias. / The increasing develpment of high performance applications is a reality on current days. However, the diversity of computer architectures, including mono and multiprocessor machines, clusters with or without front-end node, the variety of operating systems and MPI implementations has growth increasingly. Focused on this scenario, programming libraries that allows integration of several MPI implementations, operating systems and computer architectures are needed. This thesis introduces HyMPI, a MPI implementation aiming integratino, on a distributed high performance system nodes with different architectures, clusters with or without front-end machine, operating systems and MPI implementations. HyMPI offers a set of primitives based on MPI specification, including point-to-point communication, collective operations, startup and finalization and some other utility functions.

Este trabalho apresenta uma proposta de infra-estrutura de software sensível ao contexto para a computação ubíqua (ubicomp) denominada Continuum. A área de ubicomp, também chamada de computação pervasiva, pressupõe uma forte integração com o mundo real, com foco no usuário e na manutenção de alta transparência. Para o desenvolvimento de aplicativos nesse cenário, é necessária uma infra-estrutura de software adequada. A infra-estrutura projetada é baseada no padrão da arquitetura orientada a serviços (service-oriented architecture ou SOA), fazendo uso de framework e middleware, e empregando uma redefinição da semântica siga-me. Nessa visão redefinida, os usuários podem ir para qualquer lugar carregando os dados e os aplicativos que desejam, os quais podem ser usados de forma imperceptível e integrada com o mundo real (seamless integration). O foco particular desse trabalho é sensibilidade ao contexto: a percepção de características relacionadas aos usuários e ao entorno. No trabalho são considerados os recursos disponíveis no ambiente e é mantida a história dos dados de contexto. Além disso, é proposta a representação do contexto para promover raciocínio e compartilhamento de conhecimento, empregando uma ontologia. Dessa forma, contexto é representado de uma maneira formal e bastante expressiva, diferente de muitas soluções existentes hoje em dia que ainda usam modelos de representação ad hoc. Esta tese está então na interseção destas três áreas principais: infra-estrutura de software para ubicomp, sensibilidade ao contexto e ontologias. No desenvolvimento desta tese, também examina-se o campo da computação ubíqua, e sugere-se um modelo de arquitetura geral que enfrente esses desafios fundamentais. Baseado nos requisitos estabelecidos para esse modelo, propõe-se um conjunto de serviços para o Continuum. Os serviços são projetados considerando o trabalho previamente desenvolvido pelo nosso grupo de pesquisa, mais especificamente o projeto ISAM, e particularmente o middleware EXEHDA. A proposta estende esses projetos, adicionando aspectos que não haviam sido considerados no momento do seu desenvolvimento. Particularmente, o suporte a sensibilidade de contexto é melhorado com a proposta de uma ontologia para a formalização da informação de contexto. Algumas análises, usando a metodologia de estudo de caso, foram conduzidas para apreciar as principais proposições da tese. Baseado nessas avaliações, foram apresentadas algumas lições aprendidas e traçada a conclusão do trabalho. Como resultado, Continuum é uma infra-estrutura de software que endereça muitos aspectos da computação ubíqua, integrando imperceptivelmente diferentes desafios. / The present work is a proposal of a context-aware software infrastructure for ubiquitous computing (ubicomp) named Continuum. The ubicomp area, also called pervasive computing, presupposes a strong integration with the real world, with focus on the user and on keeping high transparency. For the development of applications in this scenario, we need an adequate software infrastructure. The infrastructure designed in this work is based on service-oriented architecture (SOA), making use of framework and middleware, and employing a redefinition of follow-me semantics. In this redefined vision, users can go anywhere carrying the data and application they want, which they can use in a seamlessly integrated fashion with the real world. The specific focus of our work is context awareness: the perception of characteristics related to users and surroundings. We consider the resources available in the environment and keep a history of context data. Furthermore, we propose the representation of context to promote reasoning and knowledge sharing, using ontology. In this way, context is represented in a considerably expressive, formal approach, different from many solutions that exist today, which still use ad hoc representations models. Our work is then at the intersection of these three main areas: software infrastructures for ubicomp, context awareness, and ontologies. In the development of this thesis, we also survey the field of ubiquitous computing, suggesting a general architectural model to deal with its fundamental challenges. Based on the established requirements for this model, we propose a set of services for Continuum. The services are designed considering the previous works developed by our research group, namely ISAM (Infra-estrutura de Suporte às Aplicações Móveis – Mobile Applications Support Infrastructure), and particularly the middleware EXEHDA (Execution Environment for Highly Distributed Applications). We further extend these projects, by adding aspects to them that had not been considered at the time of their development. Particularly, we improve context awareness support, proposing an ontology for the formalization of context information. We have conducted some analysis, using case study methodology, to evaluate the main propositions of our work. Based on these assessments, we present lessons learned and draw the conclusion of our work. As a result, Continuum is a software infrastructure that addresses many aspects of ubiquitous computing, seamlessly integrating many different challenges.

Algumas palavras e definições comumente utilizadas quando se está falando de Software-Defined Networking, como programabilidade, flexibilidade, ou gerenciamento centralizado, parecem muito apropriadas ao contexto de um outro paradigma de rede: Internet of Things. Em redes domésticas já não é incomum a existência de dispositivos projetados para segurança, climatização, iluminação, monitoramento de saúde e algumas formas de automação que diferem entre si em diversos aspectos, como no modo de operar e de se comunicar. Lidar com este tipo de cenário, que pode diferir bastante daquilo que estamos acostumados na gerência de redes e serviços, fazendo uso dos recursos tradicionais como ferramentas e protocolos bem estabelecidos, pode ser difícil e, em alguns casos, inviável. Com o objetivo de possibilitar o monitoramento remoto de pacientes com doenças crônicas através de dispositivos de healthcare disponíveis no mercado, uma proposta de middleware foi desenvolvida em um projeto de pesquisa para contornar as limitações relacionadas à interoperabilidade, coleta de dados, gerência, segurança e privacidade encontradas nos dispositivos utilizados. O middleware foi projetado com o intuito de executar em access points instalados na casa dos pacientes. Contudo, as limitações de hardware e software do access point utilizado refletem no desenvolvimento, pois restringem o uso de linguagens de programação e recursos que poderiam agilizar e facilitar a implementação dos módulos e dos mecanismos necessários. Os contratempos encontrados no desenvolvimento motivaram a busca por alternativas, o que resultou na refatoração do middleware através de Software-Defined Networking, baseando-se em trabalhos que exploram o uso desse paradigma em redes domésticas. O objetivo deste trabalho é verificar a viabilidade da utilização de Software-Defined Networking no contexto de Internet of Things, mais especificamente, aplicado ao serviço de monitoramento de pacientes da proposta anterior e explorar os possíveis benefícios resultantes. Com a refatoração, a maior parte da carga de serviços da rede e do monitoramento foi distribuída entre servidores remotos dedicados, com isso os desenvolvedores podem ir além das restrições do access point e fazer uso de recursos antes não disponíveis, o que potencializa um processo de desenvolvimento mais ágil e com funcionalidades mais complexas, ampliando as possibilidades do serviço. Adicionalmente, a utilização de Software-Defined Networking proporcionou a entrega de mais de um serviço através de um único access point, escalabilidade e autonomia no gerenciamento das redes e dos dispositivos e na implantação de serviços, fazendo uso de recursos do protocolo OpenFlow, e a cooperação entre dispositivos e serviços a fim de se criar uma representação digital mais ampla do ambiente monitorado. / Some words and definitions usually employed when talking about Software-Defined Networking such as programmability, frexibility, or centralized management sound very appropriate to the context of another network paradigm: Internet of Things. The presence of devices designed for security, air conditioning, lighting, health monitoring and some other automation resources have become common in home networks; those devices may be different in many ways, such as the way they operate and communicate, between others. Dealing with this kind of scenario may differ in many ways from what we are familiar regarding networking and services management; the use of traditional management tools and protocols may be hard or even unfeasible. Aiming to enable the health monitoring of patients with chronical illnesses through using off-the-shelf healthcare devices a middleware proposal was developed in a research project to circumvent interoperability, data collecting, management, security and privacy issues found in employed devices. The middleware was designed to run on access points in the homes of the patients. Although hardware and software limitations of the used access points reflect on the development process, because they restrict the use of programming languages and resources that could be employed to expedite the implementation of necessary modules and features. Development related mishaps have motivated the search for alternatives resulting in the middleware refactoring through Software-Defined Networking, based on previous works where that paradigm is used in home networks. This work aims to verify the feasability of the employment of Software- Defined Networking in the Internet of Things context, and its resulting benefits; specifically in the health monitoring of chronic patients service from the previous proposal. After refactoring most of the network and services load was distributed among remote dedicated servers allowing developers to go beyond the limitations imposed by access points constraints, and to make use of resources not available before enabling agility to the development process; it also enables the development of more complex features expanding services possibilities. Additionally Software-Defined Networking employment provides benefits such as the delivering of more than only one service through the same access point; scalability and autonomy to the network and devices monitoring, as to the service deployment through the use of OpenFlow resources; and devices and services cooperation enabling the built of a wider digital representation of the monitored environment.

En raison du grand succès des réseaux sans _l et des appareils portatifs, le paradigme de l'informatique pervasive est devenu une réalité. L'un des plus di_ciles objectifs à atteindre dans de tels environnements est de permettre à l'utilisateur d'exécuter une tâche en composant à la volée, les services et les ressources de l'environnement. Cela implique la correspondance et la sélection automatique de services à travers divers dispositifs de l'environnement pervasif. Les approches existantes considèrent souvent seulement les aspects fonctionnels des services et ne prennent pas en compte diff érents aspects non-fonctionnels tels que les préférences utilisateur, les capacités des dispositifs en termes matériels et logiciels, et l'hétérogénéité du réseau de ces dispositifs. Nous présentons une approche pour la sélection dynamique des composants et des dispositifs dans un environnement pervasif en considérant simultanément tous les aspects précédemment mentionnés. Premièrement, nous proposons une modélisation abstraite et concrète de l'application, des capacités des terminaux et des ressources, des préférences des utilisateurs, ainsi que la modélisation de la plate-forme réseau sous-jacente. Les capacit és des dispositifs sont représentées par notre extension du modèle CC/PP et les préférences des utilisateurs en utilisant notre extension du modèle CP-Net. Nous mod- élisons sous forme d'un graphe la tâche de l'utilisateur et des services réseau sous-jacent, ainsi que les exigences des services, des préférences utilisateur et les capacités des dispositifs. L'hétérogénéité des protocoles de communication est également considérée dans les graphes. Les aspects algorithmiques ont été traités en fournissant des algorithmes pour la correspondance entre les services et les composants, pour la projection des applications sur la plate-forme de composants existants et pour l'évaluation des préférences utilisateurs. Pour la description de la composition de l'application nous proposons un modèle SCA étendu. Partant d'une composition abstraite de services, nous arrivons à réaliser une composition concrète de l'application distribuée à travers les dispositifs existants. Si pendant l'exécution un nouveau meilleur dispositif apparaît, l'application est recomposée en tenant compte des nouveaux composants. Cela permet de réaliser la continuité de la session d'un dispositif vers un autre. Une mise en oeuvre d'un prototype et son évaluation sont également fournis. / Due to the large success of wireless networks and portable devices, the pervasive computing paradigm is becoming a reality. One of the most challenging objectives to be achieved in pervasive computing environments is to allow a user to perform a task by composing on the _y the environment's service and resource components. This involves automatic matching and selection of services across various devices in the pervasive environment. Existing approaches mostly consider only functional aspects for service and component matching and do not consider various non-functional aspects such as user preferences, device capabilities in terms of software and hardware, and network heterogeneity of devices. We present an approach for dynamic selection of components and devices in a pervasive environments considering all the aforementioned aspects simultaneously. First, we provide a modeling of abstract and concrete application, device capabilities and resources, user preferences as well as modeling of the underlying connected platform. Device capabilities are represented by our extended CC/PP model and user preferences using our extended CP-net model. We model both the user task and the underlying network services, along with service requirements, user preferences and device capabilities, as graphs. The heterogeneity of communication protocols is also considered in the graph. The algorithmic aspects have been treated by providing algorithms for service and component matching, application mapping on network platform and user preference evaluation. For description of application composition extended SCA model is used. Departing from an abstract composition, we arrive on achieving a concrete application composition which may be distributed across more than one device. If during the application execution a new, better device appears, the application is recomposed to replace the existing components by the newer ones. This also implies the continuity of session from one device to another. A prototype implementation and its evaluation are also provided.

Dynamic selection and composition of autonomous and loosely-coupled Web services is increasingly used to automate business processes. The typical long-running characteristic of business processes imposes new management challenges such as dynamic adaptation of running process instances. However, current process orchestration engines provide limited flexibility to dynamically adapt to changing runtime conditions (e.g., presence of faults). Additionally, current process specification languages exhibit some limitations regarding modularity of crosscutting management concerns. In particular, monitoring and adaptation logic is often scattered across several process definitions and intertwined with the business logic. This leads to monolithic and complex processes that are hard to understand, reuse, maintain, and evolve. To address these limitations, we developed a policy-based change management framework, named Manageable and Adaptable Service Compositions (MASC), to declaratively express crosscutting monitoring and process adaptation concerns in a separate and modular way. MASC policies use a set of simple, but flexible and relatively powerful, constructs to declaratively specify policies that govern: (1) discovery and selection of services to be used, (2) monitoring to detect the need for adaptation, (3) reconfiguration and adaptation of the process to handle special cases (e.g., context-dependant behaviour) and recover from typical faults in service-based processes. The identified constructs are executed by a lightweight service-oriented management middleware named MASC middleware. The adaptation is transparent because it preserves the original functional behaviour of the business process and does not tangle the adaptation logic with that of the business process. Additionally, policies do not have to be necessarily defined when designing the process; they can also be introduced later during deployment or at runtime. We implemented a MASC proof-of-concept prototype and evaluated it on Stock Trading case study scenarios. We conducted extensive studies to demonstrate the feasibility of the proposed techniques and illustrate the benefits of our approach in providing adaptive composite services using the policy-based approach. Our performance and scalability studies indicate that MASC middleware is scalable and the introduced overhead are acceptable.

The main purpose of this research is to find an effective way to personalise information searching on the Internet using middleware search agents, namely, Personalised Search Agents (PSA). The PSA acts between users and search engines, and applies new and existing techniques to mine and exploit relevant and personalised information for users. Much research has already been done in developing personalising filters, as a middleware technique which can act between user and search engines to deliver more personalised results. These personalising filters, apply one or more of the popular techniques for search result personalisation, such as the category concept, learning from user actions and using metasearch engines. By developing the PSA, these techniques have been investigated and incorporated to create an effective middleware agent for web search personalisation. In this thesis, a conceptual model for the Personalised Search Agent is developed, implemented by developing a prototype and benchmarked the prototype against existing web search practices. System development methodology which has flexible and iterative procedures that switch between conceptual design and prototype development was adopted as the research methodology. In the conceptual model of the PSA, a multi-layer client server architecture is used by applying generalisation-specialisation features. The client and the server are structurally the same, but differ in the level of generalisation and interface. The client handles personalising information regarding one user whereas the server effectively combines the personalising information of all the clients (i.e. its users) to generate a global profile. Both client and server apply the category concept where user selected URLs are mapped against categories. The PSA learns the user relevant URLs both by requesting explicit feedback and by implicitly capturing user actions (for instance the active time spent by the user on a URL). The PSA also employs a keyword-generating algorithm, and tries different combinations of words in a user search string by effectively combining them with the relevant category values. The core functionalities of the conceptual model for the PSA, were implemented in a prototype, used to test the ideas in the real word. The result was benchmarked with the results from existing search engines to determine the efficiency of the PSA over conventional searching. A comparison of the test results revealed that the PSA is more effective and efficient in finding relevant and personalised results for individual users and possesses a unique user sense rather than the general user sense of traditional search engines. The PSA, is a novel architecture and contributes to the domain of knowledge web information searching, by delivering new ideas such as active time based user relevancy calculations, automatic generation of sensible search keyword combinations and the implementation of a multi-layer agent architecture. Moreover, the PSA has high potential for future extensions as well. Because it captures highly personalised data, data mining techniques which employ case-based reasoning make the PSA a more responsive, more accurate and more effective tool for personalised information searching.

Mobility in distributed object systems is useful as it can provide such properties as load balancing, code to data movement, fault tolerance, migration to stable storage, and autonomous semantics. In a widely distributed system, these properties are important as they can help alleviate latency issues and increase performance within the system. Additionally, they provid more flexibility in the programming of distributed systems by relaxing static location restrictions. Location transparency removes the need for client objects to explicitly know or define the location of a server object when communicating. If a server object is capable of migration, relocation transparency maintains reference validity throughout the migration. Several models for providing relocation transparency exist, including the home location, forwarding location, and broadcast models. This thesis proposes a model that uses a distributed registry system and dynamic reference updating to provide location and relocation transparency. A registry system is used to provide location independence by resolving a location independent name to a reference that can be used by a client. A naming system is used to provide correct binding and production of names within the naming restrictions of the system. The thesis proposes that the choice of naming system within a distributed or mobile object system has a large effect on the system's ability to support efficient transparent object relocation. This thesis proposes that a formal analysis of naming systems enables the selection of an appropriate naming system for a distributed or mobile object system given the object system's naming, distribution and transparency requirements. This thesis presents a new classification scheme for naming systems, based on analysis of a broad spectrum of naming systems. A classification of existing mobile and distributed object systems with respect to existing naming models is provided. It is shown that the current models need to be refined and extended to completely and correctly classify the example systems. This thesis proposes extensions and refinements that enable correct and complete classification of mobile and distributed object systems with a need for transparency. The extended naming model is then used to describe a naming system that is capable of implementing any naming system classifiable by the extended model. A classification of a naming system to support the proposed model of location and relocation transparency is presented. A distributed ORB system is designed and implemented to support the distributed namespace and generic naming system implementation. The distributed ORB system is hierarchically structured and is capable of adapting in response to node failure. This ORB system is used to support client and server object integration in the DISCWorld metacomputing environment. The ORB system is used to provide migration, replication and cloning services to the DISCWorld metacomputing environment. A qualitative analysis of the generic naming system and the DISCWorld ORB system is performed. A comparison between the proposed model for location and relocation transparency and existing models is also presented. This comparison shows that the proposed model exhibits better location and relocation performance within the DISCWorld environment. The distributed nature of the ORB system and namespace provides a scalable nature in terms of namespace size, the number of objects within the system, and the frequency of location and relocation requests. / Thesis (Ph.D.)--Engineering (Department of Computer Science), 2000.

Reconfiguration dynamique des architectures orientées services

This thesis aims at two different goals. First, a uniform presentation of the major relevant research results on EAOP-based expressive aspects. We motivate that these instantiations enable aspects to be defined more concisely and provide better support for formal reasoning over AO programs than standard atomic approaches and other proposed non-atomic approaches. Concretely, four groups of results are presented in order to substantiate these claims: 1. The EAOP model, which features pointcuts defined over the execution history of an underly- ing base program. We present a taxonomy of the major language design issues pertaining to non-atomic aspect languages, such as pointcut expressiveness (e.g., finite-state based, turing- complete) and aspect composition mechanisms (e.g., precedence specifications vs. turing- complete composition programs). 2. Support for the formal definition of aspect-oriented programming based on different seman- tic paradigms (among others, operational semantics and denotation semantics). Furthermore, we have investigated the static analysis of interactions among aspects as well as applicability conditions for aspects. The corresponding foundational work on AOP has also permitted to investigate different weaver definitions that generalize on those used in other approaches. 3. Several instantiations of the EAOP model for aspects concerning sequential program execu- tions, in particular, for component-based and system-level programming. The former has re- sulted in formally-defined notions of aspects for the modification of component protocols, while the latter has shown, in particular that expressive aspects can be implemented in a performance- critical domain with negligible to reasonable overhead. 4. Two instantiations of the EAOP model to distributed and concurrent programming that signifi- cantly increase the abstraction level of aspect definitions by means of domain-specific abstrac- tions.