A Multi-Criteria Framework to Assist on the Design of Internet-of-Things Systems

• Main Author: Silva, Edgar Miguel Felício Oliveira da
• Publication Date: 2020
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10362/105227
• topic: Internet-of-Things; Resource-Constrained Systems; Model-Driven Engineering; Interoperability; Multi-Criteria Decision; IoT Systems Assessment; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

The Internet-of-Things (IoT), considered as Internet first real evolution, has become immensely important to society due to revolutionary business models with the potential to radically improve Human life. Manufacturers are engaged in developing embedded systems (IoT Systems) for different purposes to address this new variety of application domains and services. With the capability to agilely respond to a very dynamic market offer of IoT Systems, the design phase of IoT ecosystems can be enhanced. However, select the more suitable IoT System for a certain task is currently based on stakeholder’s knowledge, normally from lived experience or intuition, although it does not mean that a proper decision is being made. Furthermore, the lack of methods to formally describe IoT Systems characteristics, capable of being automatically used by methods is also an issue, reinforced by the growth of available information directly connected to Internet spread. Contributing to improve IoT Ecosystems design phase, this PhD work proposes a framework capable of fully characterise an IoT System and assist stakeholder’s on the decision of which is the proper IoT System for a specific task. This enables decision-makers to perform a better reasoning and more aware analysis of diverse and very often contradicting criteria. It is also intended to provide methods to integrate energy consumptionsimulation tools and address interoperability with standards, methods or systems within the IoT scope. This is addressed using a model-driven based framework supporting a high openness level to use different software languages and decision methods, but also for interoperability with other systems, tools and methods.