Enhanced Hybrid Equilibrium Strategy in Fog-Cloud Computing Networks with Optimal Task Scheduling
作者机构:School of Computer ScienceYangtze UniversityJingzhou434000China
出 版 物:《Computers, Materials & Continua》 (计算机、材料和连续体(英文))
年 卷 期:2024年第79卷第5期
页 面:2647-2672页
核心收录:
学科分类:08[工学] 0805[工学-材料科学与工程(可授工学、理学学位)] 081201[工学-计算机系统结构] 0812[工学-计算机科学与技术(可授工学、理学学位)]
基 金:in part by the Hubei Natural Science and Research Project under Grant 2020418 in part by the 2021 Light of Taihu Science and Technology Project in part by the 2022 Wuxi Science and Technology Innovation and Entrepreneurship Program
主 题:Artificial intelligence of things fog computing task scheduling equilibrium optimizer differential evaluation algorithm local search
摘 要:More devices in the Intelligent Internet of Things(AIoT)result in an increased number of tasks that require low latency and real-time responsiveness,leading to an increased demand for computational *** computing’s low-latency performance issues in AIoT scenarios have led researchers to explore fog computing as a complementary ***,the effective allocation of resources for task execution within fog environments,characterized by limitations and heterogeneity in computational resources,remains a formidable *** tackle this challenge,in this study,we integrate fog computing and cloud *** begin by establishing a fog-cloud environment framework,followed by the formulation of a mathematical model for task ***,we introduce an enhanced hybrid Equilibrium Optimizer(EHEO)tailored for AIoT task *** overarching objective is to decrease both the makespan and energy consumption of the fog-cloud system while accounting for task *** proposed EHEO method undergoes a thorough evaluation against multiple benchmark algorithms,encompassing metrics likemakespan,total energy consumption,success rate,and average waiting *** experimental results unequivocally demonstrate the superior performance of EHEO across all assessed ***,in the most favorable conditions,EHEO significantly diminishes both the makespan and energy consumption by approximately 50%and 35.5%,respectively,compared to the secondbest performing approach,which affirms its efficacy in advancing the efficiency of AIoT task scheduling within fog-cloud networks.