Abstract
Today's telecommunication market is populated by a great variety of available networking solutions. Given the exponential growth of network infrastructures and customers' needs, increasing system diversity is inevitable. A key emerging trend is the convergence of heterogeneous personal devices into spontaneous infrastructure-less networks. In such a diverse environment, flexible middleware technologies can serve as the "glue" between heterogeneous systems. Products from different vendors and network devices with different architectures could be brought together using a "common language" offered by middleware. The objective of this thesis is to investigate middleware technologies as a vehicle of managing emerging, fixed and mobile networks. Four different middleware technologies are compared and evaluated in terms of performance and usability. The investigated technologies include the well- established CORBA platform. Mobile Agents and the XML-based Web Services/SOAP and XML-RPC technologies. XML-based technologies present a number of attractive features, such as the use of the HTTP transport protocol and easy integration with XML-structured data and Web browsers. On the other hand. Mobile Agents offer the inherent feature of software migration that can be used for network element programmability and capability enhancement, albeit at a high performance cost. Two different middleware case studies are examined in order to assess their suitability for emerging network architectures. The first case study investigates an Agent-based middleware system for managing Quality of service in IP Differentiated Services networks. The middleware platform is addressed to a fixed network infrastructure, demonstrating the integration of heterogeneous network elements. The second case study investigates a middleware-based programmable infrastructure that allows the nodes of a mobile ad hoc network to download and activate required protocol and service software dynamically. This enables the alignment of nodes capabilities and allows, for instance, full-scale quality of service-based communication among heterogeneous ad hoc network nodes.