This thesis work takes an investigative approach into developing a middleware solution for managing services in a community cloud computing infrastructure predominantly made of interconnected low power wireless devices. The thesis extends itself slightly outside of this acute framing to ensure heterogeneity is accounted for. The developed framework, in its draft implementation, provides networks with value added functionality in a way which minimally impacts nodes on the network. Two sub-protocols are developed and successfully implemented in order to achieve efficient discovery and allocation of the community cloud resources. First results are promising as the systems developed show both low resource consumption in its application, but also the ability to effectively transfer services through the network while evenly distributing load amongst computing resources on the network.

Memoria para optar al título de Ingeniero Civil Eléctrico / El paradigma de comunicación Internet of Things (IoT), el cual plantea la posibilidad de interconectar objetos cotidianos y toda clase de dispositivos convencionales a Internet, está actualmente en pleno desarrollo. El gran número de nodos que se espera conectar a Internet exige a su vez la implementación a gran escala de Internet Protocol versión 6 (IPv6). IoT busca el desarrollo de nuevas aplicaciones y ha impulsado la creación de nuevas arquitecturas de red y nuevas clases de dispositivos. Las redes Low Power Wide Area Networks (LPWAN) han surgido recientemente como una evolución natural del concepto Wireless Sensor Networks (WSN), redes de sensores in- terconectadas. A la luz del IoT, las nuevas redes LPWAN abren un nuevo campo de desarrollo, principalmente enfocado en servicios de monitoreo y afines que se desarrollen en áreas am- plias y no requieran grandes tasas de transferencia. Los dispositivos LPWAN se caracterizan por ser de bajo consumo energético y de bajo costo, facilitando su despliegue masivo por largos períodos sin necesidad de recargar sus baterías. Long Range Wide Area Network (LoRaWAN) es una de las primeras y principales tec- nologías LPWAN, y presenta una gran flexibilidad que la hace ideal para redes de diseño propio. En América funciona en la banda industrial, científica y médica (ISM) alrededor de los 915 MHz. Sin embargo, también existen muchas otras tecnologías LPWAN con arquitec- turas y protocolos propietarios, lo que dificulta alcanzar la interoperabilidad que se desea en el entorno IoT. El grupo de trabajo para la implementación de IPv6 sobre redes LPWAN (lpwan WG) perteneciente al Internet Engineering Task Force (IETF) se encuentra actualmente desarrol- lando un mecanismo de compresión y fragmentación de paquetes IPv6 para redes LPWAN denominado Static Context Header Compression (SCHC). El esquema de compresión se en- cuentra terminado, pero aún no ha sido implementado ni evaluado de manera oficial. En este trabajo se presenta una plataforma experimental para la implementación y eval- uación del mecanismo SCHC sobre una red LoRaWAN consistente en un nodo terminal Mi- crochip y un Radio Gateway (RG) de Everynet. En su desarrollo se han integrado múltiples y diversas herramientas del campo de las Telecomunicaciones y las Tecnologías de Información y Comunicación (ICT). La plataforma creada logra una implementación básica pero exitosa del esquema de com- presión de SCHC. Por medio de ella se ha llevado a cabo una evaluación preliminar del funcionamiento de SCHC, analizando el nivel de compresión logrado por el mecanismo para tres contextos de comunicación característicos de una red LPWAN. Los resultados obtenidos son positivos.

The growth of the World Wide Web has in large part been made possible by the technologies that power it. These are the Web servers and Web browsers that many take for granted. Each has evolved to support the changing needs of users of the WWW from simple static text to highly interactive and dynamic multimedia. The Web servers, in particular, have evolved into a spectrum of different technologies to support what are now known as Web applications. These are usually installed and accessed through a Web server. Security is a problem in Web server environments and therefore the Web servers are usually run as an un-privileged user. Performance is another problem as some of these technologies require re-initialization of the execution environment with every subsequent request. These security and performance shortcomings have been dealt with by numerous Web application technologies. Most of these technologies are language-centric and seek solutions to the security and performance shortcomings independently of each other. The universal Web application server is proposed as an alternative solution addressing the security, language dependence and performance shortcomings of existing technologies. It has support for multiple authors in a secure environment with support for multiple implementation technologies (languages) using persistent interpreters to enhance performance. Test results from the performance evaluation show that the introduction of the layers of processing contributes a small percentage to the total request processing time and that the universal Web application server can perform comparably to other Web application servers. Tests with twenty users also showed that packaging and deploying Web applications in the universal Web application server is an easy and viable approach. Moreover, the installation of PhpBB2 in the universal Web application server demonstrates that it can be used with realistic Web applications. A universal Web application server that provides an efficient, secure and language independent environment has been developed and thoroughly evaluated demonstrating that a Web application server that addresses the shortcomings of existing technologies is feasible.

Computer scientists must align social and technical factors for communication technologies in developing regions yet lack a framework to do so. The novel Softbridge framework comprises several components to address this gap. The Softbridge stack abstraction supplements the established Open Systems Interconnect model with a collection of technical layers clustered around 'people' issues. The Softbridge stack aligns the technological design of communication systems with awareness of social factors characteristic of developing regions. In a similar fashion, a new evaluation abstraction called Quality of Communication augments traditional Quality of Service by considering socio-cultural factors of a user's perception of system performance. The conceptualisation of these new abstractions was driven by long-term experimental interventions within two South African digital divides. One field study concerned communication bridges for socio-economically disadvantaged Deaf users. The second field study concerned a wireless telehealth system between rural nurses and doctors. The application domains were quite different yet yielded similarities that informed the Softbridge and Quality of Communication abstractions. The third Softbridge component is an iterative socially aware software engineering method that includes action research. This method was used to guide cyclical interventions with target communities to solve community problems with communication technologies. The Softbridge framework components are recursive products of this iterative approach, emerging via critical reflection on the design, evaluation and methodological processes of the respective field studies. Quantitative and qualitative data were triangulated on a series of communication prototypes for each field study with usage metrics, semi-structured interviews, focus groups and observation in the field. Action research journals documented the overall process to achieve post-positivist recoverability rather than positivistic replicability. Analysis of the results from both field studies was iteratively synthesised to develop the Softbridge framework and consider its implications. The most significant finding is that awareness of social issues helps explain why users might not accept a technically sound communication system. It was found that when facilitated effectively by intermediaries, the Softbridge framework enables unintended uses of experimental artefacts that empower users to appropriate communication technologies on their own. Thus, the Softbridge framework helps to align technical and socio-cultural factors.

Bibliography: leaves 272-280. / xi, 280 leaves : col. ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.)--University of Adelaide, Dept. of Computer Science, 2000

Thesis (Ph.D.) -- University of Adelaide, Dept. of Computer Science, 2000. / Bibliography: leaves 272-280.

Thesis (Ph. D.)--University of Washington, 2007. / Vita. Includes bibliographical references (leaves 113-133).

Magister Scientiae - MSc / This thesis shows that mobility management protocols for infrastructure Internet may be used in a wireless mesh network environment. In this research Mobile IPv6 and Fast Handover for Hierarchical Mobile IPv6 are successfully implemented in a wireless mesh network environment. Two experiments were carried out: vertical and horizontal handover simulations. Vertical handover simulation involved a heterogeneous wireless environment comprising both wireless local area and wireless mesh networks. An OPNET Mobile IPv6 model was used to simulate the vertical handover experiment. Horizontal handover simulation involved Mobile IPv6 and Fast Handover for Hierarchical Mobile IPv6 applied in ns2 wireless mesh network. The vertical handover results show that MIPv6 is able to manage vertical handover between wireless local area and wireless mesh network. The horizontal handover results illustrate that in mesh networks, Fast Handover for Hierarchical Mobile IPv6's performance is superior to Mobile IPv6. Fast Handover for Hierarchical Mobile IPv6 generates more throughput and less delay than Mobile IPv6. Furthermore, Fast Handover for Hierarchical Mobile IPv6 drops less data packets than Mobile IPv6. The simulations indicate that even though there are multi-hop communications in wireless mesh networks, the performance of the multi-hop routing may not play a big role in the handover performance. This is so because the mesh routers are mostly static and the multi-hop routes are readily available. Thus, the total handover delay is not affected too much by the WMN hops in the paths for signaling message transmission. / South Africa

No description available.

No description available.