A modular architecture for the emulation of scalable edge applications

• Ano de defesa: 2024
• Autor(a) principal: Silva, Matheus Lyra da
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul; Escola Politécnica; Brasil; PUCRS; Programa de Pós-Graduação em Ciência da Computação
• Palavras-chave em Português: Edge Computing; Internet of Things; Industry 4.0; Emulation Architecture; Computação de Borda; Internet das Coisas; Indústria 4.0; Arquitetura de Emulação; CIENCIA DA COMPUTACAO::TEORIA DA COMPUTACAO
• Link de acesso: https://tede2.pucrs.br/tede2/handle/tede/11679

In the last decade, cloud computing transformed the information technology industry, allowing companies to execute many services that require on-demand availability of computational resources with more flexible provisioning and cost models, including the processing of already growing data volumes. But in the past few years, other technologies such as internet of things and the digitized industry known as industry 4.0 have emerged, increasing data generation even more. The large amounts of data produced by user devices and manufacturing machinery have made both industry and academia search for new approaches to process all this data. Alternatives to the cloud centralized processing model and its inherent high latencies have been studied and edge computing is being proposed as a solution to these problems. In this context, to design efficient solutions, it is necessary to consider which application components must be allocated at the edge to effectively improve performance and reduce the cost of execution, which raises the urge for tools that enable the design and experimentation over specific edge computing scenarios. In this thesis, we contribute to the state-of-the-art of edge computing and explore the possibilities of emulating edge scenarios considering scalability aspects and different infrastructures, allowing experimentation and design with realistic results. We propose a modular emulation architecture for prototyping scalable edge applications. The proposed architecture is adaptable to many different conditions. We enable measurements of specific scenarios considering the computational footprint of edge devices and explore the possibilities of evaluating different behaviors of edge applications. Furthermore, we validate a prototype with a smart agriculture use-case, in which the results of the scalability experimentation pointed 200 as the observed maximum number of Edge Devices that a single Edge Broker instance can support on the applied restrictions and testbed configurations running this implementation.