Exploring Factors Impacting Data Offloading Performance in Edge and Cloud Environments

Recently, the deployments of smart IoT devices, such as wearable devices and drones, have risen at an unprecedented pace. But these edge devices typically have limited computing and storage resources, thus making it hard to perform compute-intensive jobs on their acquired data despite growing demands. In this paper, we uncover five major factors and present our empirical evaluation results in an offloading environment utilizing external servers. From the findings, we present a taxonomy of factors that should be considered along with different data types for offloading the data of edge devices to external nodes more efficiently.