Elastic Data Analytics for the Cloud-to-Things Continuum

Abstract

The massive deployment of Internet-connected devices has led to an increase in the collection of data that are then used by companies to improve their decision-making processes. This growing trend demands more and more cloud and communications infrastructure. The limited resources, the need for sharing them, and the fact that many consumers are interested in the same data, call for an efficient management of the available resources. The Cloud-to-things continuum can be used to execute different analytics closer to the data source so that infrastructure consumption and data circulation can be optimized. In this article, we propose using elastic data analytics, whose behaviour can be dynamically modified according to the quality requirements defined by each IoT system and the available resources of the Cloud-to-things infrastructure. In addition, a framework for dynamically modifying their behavior is proposed.