A Multi-Classifiers Based Algorithm for Energy Efficient Tasks Offloading in Fog Computing

Moteb K. Alasmari

doi:10.3390/s23167209

Return

A Multi-Classifiers Based Algorithm for Energy Efficient Tasks Offloading in Fog Computing

2023-08-16

Moteb K. Alasmari *

Sami S. Alwakeel

Yousef Al-Ohali

DOI：10.3390/s23167209

Original

Save

View PDF

Abstract

En 中文

The IoT has connected a vast number of devices on a massive internet scale. With the rapid increase in devices and data, offloading tasks from IoT devices to remote Cloud data centers becomes unproductive and costly. Optimizing energy consumption in IoT devices while meeting deadlines and data constraints is challenging. Fog Computing aids efficient IoT task processing with proximity to nodes and lower service delay. Cloud task offloading occurs frequently due to Fog Computing's limited resources compared to remote Cloud, necessitating improved techniques for accurate categorization and distribution of IoT device task offloading in a hybrid IoT, Fog, and Cloud paradigm. This article explores relevant offloading strategies in Fog Computing and proposes MCEETO, an intelligent energy-aware allocation strategy, utilizing a multi-classifier-based algorithm for efficient task offloading by selecting optimal Fog Devices (FDs) for module placement. MCEETO decision parameters include task attributes, Fog node characteristics, network latency, and bandwidth. The method is evaluated using the iFogSim simulator and compared with edge-ward and Cloud-only strategies. The proposed solution is more energy-efficient, saving around 11.36% compared to Cloud-only and approximately 9.30% compared to the edge-ward strategy. Additionally, the MCEETO algorithm achieved a 67% and 96% reduction in network usage compared to both strategies.

Keywords:

fog computing

internet of things

module placement

energy efficiency

task offloading

iFogSim

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.