[Zurück]

S. Dustdar, I. Murturi:
"Towards Distributed Edge-based Systems";
Vortrag: IEEE 2nd International Conference on Cognitive Machine Intelligence (CogMI 2020) - Online Conference, Atlanta, Georgia, USA; 01.12.2020 - 03.12.2020; in: "Proceedings of the IEEE 2nd International Conference on Cognitive Machine Intelligence (CogMI 2020)", IEEE, (2020), ISBN: 978-1-7281-4145-9; S. 92 - 100.

In the past few years, researchers from academia and industry stakeholders suggest adding more computational resources (i.e., storage, networking, and processing) closer to the end-users and IoT domain, respectively, at the edge of the network. Such computation entities perceived as edge devices aim to overcome high-latency issues between the cloud and the IoT domain. Thus, processing IoT data streams closer to the end-users and IoT domain can solve several operational challenges. Since then, a plethora of application-specific IoT systems are introduced, mainly hard-coded, inflexible, and limited extensibility for future changes. Additionally, most IoT systems maintain a centralized design to operate without considering the dynamic nature of edge networks. In this paper, we discuss some of the research issues, challenges, and potential solutions to enable: i) deploying edge functions on edge resources in a distributed manner and ii) deploying and scaling edge applications on-premises of Edge-Cloud infrastructure. Additionally, we discuss in detail the three-tier Edge-Cloud architecture. Finally, we introduce a conceptual framework that aims to enable easy configuration and deployment of edge-based systems on top of heterogeneous edge infrastructure and present our vision within a smart city scenario.

Edge-Cloud Continuum, Edge-based Systems, Distributed Edge Function

http://dx.doi.org/10.1109/CogMI50398.2020.00021

Projektleitung Schahram Dustdar:
FogProtect