Design of Platform-Independent IoT Applications in the Edge-Cloud Continuum

In the dynamic edge cloud continuum, where platforms and vendors do not have shared standards, application development often depends on the initial platform choice. Furthermore, applications designed and deployed to leverage the cloud for data analytics and storage, with devices as data generators, face challenges in adapting to intermediate levels of the edge-cloud continuum (such as on-premise, near and far-edge). This paper presents a novel approach that uses a set of abstractions to conceptualize platform-independent applications. Applications are modeled as a composition of abstract services, which are independent of the place of distribution, be it edge or cloud, and specific platforms, and offer developers flexibility in the selection of services and providers during distribution, optimizing the use of resources and costs. Through detailed modeling, this paper uncovers how applications interact with abstract services, structured as workflows comprising interconnected tasks encompassing computation, communication, and storage. The interaction between these elements has a significant impact on both quality of service (QoS) and cost considerations, a key consideration in the design phase. Using such abstractions in modeling not only facilitates application development but also simplifies application deployment processes. A case study within the Industrial Internet of Things (IIoT) validates the modeling approach, illustrating how precise and abstract modeling of the application, independent of specific platforms, facilitates the development of real-world applications within the edge-cloud continuum.