Fog Horizons -- A Theoretical Concept to Enable Dynamic Fog Architectures

Abstract

With the Internet of Things, more and more devices, and therefore data, need to be handled by current architectures. The flood of provided information leads to bottlenecks within regular client-server architectures. In order to address this challenge and to enable new applications, fog computing is a promising architectural style which pushes the data and device handling closer to the edge. Thus, systems are getting more decentralized and -- with the inclusion of devices with constantly changing positions such as cars, drones and smartphones -- more dynamic. To address the corresponding challenges, we defined four concepts for dynamic Fog Architectures: Fog Visibility, Fog Horizon, Fog Reachability, and finally, a redefinition of the Fog Architecture, based on the previous concepts. These concepts support a common understanding of the constantly changing sets of components these fog environments have to deal with. In addition to the common understanding, the concepts are also helpful to set up architectures, discover communication partners as well as setting scopes. We provide mathematical definitions for each concept and give examples. In the case study, we applied these concepts in four applications from different domains to show the applicability and generalizability.