Towards Semantic Modelling of the Edge-Cloud Continuum

Abstract

Fog and edge computing paradigms increase the performance of IoT systems compared to those based exclusively on conventional cloud computing. Basically, they propose to move software services that process information to a nearby place where IoT data is collected instead of at the core of the network. The computing continuum concept goes a step further and proposes to run such software services in a transparent manner at any of the different computing paradigms and, if the execution context changes (for example due to unforeseen contingencies), to move the services to other devices if they may increase performance. In this context, advanced mechanisms are required to successfully transfer those software services to devices hosted in different computing platforms located anywhere from the edge to the cloud. This article proposes the Edge Cloud Computing ontology (ECO), an ontology for IoT systems composed of devices and data centers hosted in edge, fog or cloud computing environments. We also expose an example scenario based on a service architecture on which ECO facilitates management actions. These actions include detecting the overload status of system elements, proposals for suitable locations for software deployment or identifying elements potentially affected by problems in the system, such as connection link failures.