On-demand deployment for IoT applications
Jingbin Zhang,Meng Ma,Wei He,Ping Wang
DOI: https://doi.org/10.1016/j.sysarc.2020.101794
IF: 5.836
2020-12-01
Journal of Systems Architecture
Abstract:<p>The uncertainties of IoT-edges system and environment make challenges to applications in reliability, stability, and total latency, etc. With edge devices increasingly able to connect to cloud servers from anywhere, applications can potentially perform deployment changing at runtime to improve performance. In this paper, we propose a dynamic-programming based algorithm, named <em>E-ODD</em>, for the on-demand deployment of applications at runtime. When the device does not have enough computational resources, it offloads some movable tasks to the cloud to occupy additional resources. However, when it is difficult to occupy cloud resources or computational resources on the device become excessive, the device migrates tasks back. Besides, we present MODE, a generic and flexible middleware for edge cloud on-demand deployment. We propose a context model and detect corresponding complex events which trigger the deployment changes. Based on the event detection model, middleware can acquire changing information real-timely. We have successfully applied our middleware for ventricular fibrillation monitoring. Finally, experiments prove that our on-demand deployment model outperforms other selected models both in total latency and throughput, especially in dynamic environments.</p>
computer science, software engineering, hardware & architecture