期刊
IEEE TRANSACTIONS ON CLOUD COMPUTING
卷 11, 期 2, 页码 2144-2157出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCC.2022.3188926
关键词
Task analysis; Internet of Things; Cloud computing; Processor scheduling; Energy consumption; Statistics; Sociology; Cloud computing; fog computing; Internet of Thing (IoT); multi-objective optimization; non-dominated sorting; scheduling
In recent years, the development of Internet of Things (IoT) has made it one of the most important technologies. The fog computing architecture has partially addressed the issue of latency and other limitations of the IoT-based cloud computing paradigm, but an appropriate and efficient task scheduling method considering energy consumption is still needed. This article proposes a constraint bi-objective optimization problem and a directed non-dominated sorting genetic algorithm (D-NSGA-II) to minimize servers' energy consumption and overall response time simultaneously. Experimental results show that D-NSGA-II outperforms other algorithms and can meet all request deadlines.
In recent years, the Internet of Things (IoT) developments have made it one of the most important technologies. The exponential growth of data and increasing the number of latency-sensitive applications has necessitated a new approach to support these applications. The emerging fog computing architecture has partially addressed the issue of latency and other limitations of the IoT-based cloud computing paradigm. In order to achieve high-quality services and high system performance, an appropriate and efficient task scheduling method is needed, in addition, the energy consumption of computing devices should be considered. In this article, a constraint bi-objective optimization problem is designed to minimize the servers' energy consumption and overall response time simultaneously. Then, to solve this problem, by introducing a recombination operator and modifying NSGA-II, a directed non-dominated sorting genetic algorithm, called D-NSGA-II is proposed. This algorithm can control the selection pressure of agents, and balance the exploration and exploitation abilities of the algorithm using this new operator. To evaluate the performance of this algorithm, it is compared with well-known meta-heuristic algorithms. The experimental results demonstrate the D-NSGA-II has better performance than other algorithms. It can also respond to all requests before their deadline.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据