Robustness via Elasticity Accelerators for the IoT-Edge-Cloud Continuum
Hong‐Linh Truong, Kostas Magoutis · 2022
This paper presents a novel framework to enhance programmability of the IoT-edge-cloud continuum and to accommodate rapid change through software composition and elastic adaptation using appropriate reusable runtime components, techniques, and languages (which we collectively term Accelerators) to optimize resources and services. We particularly pinpoint elasticity as a driver for increased robustness, an ability we abbreviate as $\mathrm{R}_{\mathrm{V}}$E. $\mathrm{R}_{\mathrm{V}}$E Accelerators aim at enabling end-to-end programming, configuration, monitoring, and optimization in the IoT-edge-cloud continuum for emerging swarm applications and services, with the ability to coordinate domain-specific elasticity capabilities in a cross-layered manner. $\mathrm{R}_{\mathrm{V}}\mathrm{E}$ Accelerators will enable developers to easily program adaptive edge-cloud functionality through different layers, while cloud and edge solution providers can coordinate policies across layers to support robust, resilient systems at low effort and cost for their customers.