Development and Analysis of Middleware for Fog Computing On Energy Efficient Devices Suiting the Internet of Things

The emergence of next-generation infrastructures is facilitating the transition from centralized cloud computing to the cloud-to-fog continuum, wherein services are disseminated over many layers, ranging from centralized clouds to decentralized edge devices. This model facilitates more efficient and responsive service provisioning, especially in latency-sensitive and resource-limited contexts. In these situations, middleware is essential for managing the distributed characteristics of services, aiding energy-efficient devices at the network periphery, and enabling seamless service orchestration. These frameworks facilitate the monitoring of small-scale services and nodes, hence maintaining performance consistency. Subsequently, distributions can be refined by integrating insights from real-time measurements and network topology data, facilitating dynamic resource allocation and improving system efficiency. Additionally, emulation frameworks are essential for testing diverse networking scenarios and finding nearly perfect service deployments in a sandboxed environment. By determining the most appropriate emulator for application development, we provide an easy-to-use platform to researchers and developers, such that they can evaluate their services' needs. Moreover, novel network paradigms and protocols are easily adaptable on given emulations. This dissertation examines the design of middleware solutions within the cloud-to-fog continuum and analyzes the application of emulation tools to enhance service delivery at the network edge.