Semantic Model of Variability and Capabilities of IoT Applications for Embedded Software Ecosystems

Applications in embedded open software ecosystems for Internet of Things devices open new challenges regarding how their variability and capabilities should be modeled. In collaboration with an industrial partner, we have recognized that such applications have complex constraints on the context. We have also identified a need to model their deployment topology and functionality in order to enable their seamless integration into the platform. In this paper, we draw from research in related fields and present a model of IoT applications. It is built using semantic annotations and uses semantic reasoning to resolve context requirements. We present the implications on the architecture of the ecosystem and the concepts defined in the model. Finally, we discuss the evaluation of the model and its benefits and liabilities. Although the approach results in more complex descriptions of applications, we conclude that it is suitable for modeling applications in IoT software ecosystems since it is more adaptable and expressive than the alternatives.