학술논문

Decentralized Adaptive Latency-Aware Cloud-Edge-Dew Architecture for Unreliable Network
Document Type
Conference
Source
Proceedings of the 2019 11th International Conference on Machine Learning and Computing. :142-146
Subject
Cloud computing
cloud-dew architecture
computational offloading
dew computing
distributed computing
edge computing
Language
English
Abstract
Smart end-user devices are connected to the global ecosystem explosively and producing an enormous amount of network traffic at the backhaul. Moreover, Real-time applications such as remote surgery, self-driving cars, and other new technologies required high quality of user experience. To address the challenges Cloud Computing is extended to a new paradigm known as Dew Computing which brings cloud services and capabilities closer to end user devices based on proximity through a decentralized exchange of data and information. However, there is still a user requirement for Ultra-low latency and reliability so that, we introduced Cloud-Edge-Dew architecture to form adaptive local resource utilization and computational offloading during unreliable network to facilitate the collaboration between the various layer in the hierarchy. Moreover, smart end-user devices establish a peer communication or accessing the micro-services which are delivered from Dew Servers and Edge Server. As a result, our scheme provides a decentralize local computation which is more efficient in response time, availability and storage.

Online Access