View in EDS

GWO based energy-efficient workflow scheduling for heterogeneous computing systems.

Saved in:
Bibliographic Details
Title: GWO based energy-efficient workflow scheduling for heterogeneous computing systems.
Authors: Karishma1 (AUTHOR) maths.karishma97@gmail.com, Kumar, Harendra1 (AUTHOR) balyan.kumar@gmail.com
Source: Soft Computing - A Fusion of Foundations, Methodologies & Applications. Apr2025, Vol. 29 Issue 7, p3469-3508. 40p.
Subjects: Computational mathematics, Fast Fourier transforms, Gaussian elimination, Artificial intelligence, Energy consumption, Heterogeneous computing, Metaheuristic algorithms
Abstract: This article addresses the critical challenge of energy-efficient task scheduling in heterogeneous computing systems, which are known for their superior performance and complexity management across diverse applications. The research proposes two novel task scheduling algorithms based on the metaheuristic grey wolf optimization technique to optimize energy consumption while minimizing computational time for parallel applications. In the first algorithm, the primary objective is to mitigate static energy consumption while simultaneously enhancing computational efficiency. This is achieved by grouping the most energy-efficient tasks into clusters. While the second algorithm employs a refined approach to minimize dynamic energy consumption. This involves the utilization of a dynamic voltage and frequency scaling-enabled grey wolf optimization model, meticulously crafted for the allocation of these task clusters onto the most suitable processors. The proposed algorithms are rigorously evaluated using real-world applications, including fast Fourier transform, Gaussian elimination, and randomly generated parallel applications. The experimental findings affirm that the proposed approach consistently improves performance across various metrics, including energy consumption, degree of imbalance, resource utilization, sensitivity analysis and task assignment computation time. Specifically, it significantly reduces computational time, achieving a reduction of 30–47% for fast Fourier transform applications and 44–48% for Gaussian elimination applications compared to existing algorithms. Additionally, it improves resource utilization, with enhancements ranging from 12.28 to 45.99% when compared to a variety of existing algorithms. To further validate the effectiveness of the proposed approach, it is applied to 10 real-world optimization problems. The performance is benchmarked against three state-of-the-art algorithms from the 'CEC2020 Competition': SASS, sCMAgES, and COLSHADE. This comprehensive evaluation framework provides a robust assessment of the algorithms' efficacy. [ABSTRACT FROM AUTHOR]
Copyright of Soft Computing - A Fusion of Foundations, Methodologies & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:This article addresses the critical challenge of energy-efficient task scheduling in heterogeneous computing systems, which are known for their superior performance and complexity management across diverse applications. The research proposes two novel task scheduling algorithms based on the metaheuristic grey wolf optimization technique to optimize energy consumption while minimizing computational time for parallel applications. In the first algorithm, the primary objective is to mitigate static energy consumption while simultaneously enhancing computational efficiency. This is achieved by grouping the most energy-efficient tasks into clusters. While the second algorithm employs a refined approach to minimize dynamic energy consumption. This involves the utilization of a dynamic voltage and frequency scaling-enabled grey wolf optimization model, meticulously crafted for the allocation of these task clusters onto the most suitable processors. The proposed algorithms are rigorously evaluated using real-world applications, including fast Fourier transform, Gaussian elimination, and randomly generated parallel applications. The experimental findings affirm that the proposed approach consistently improves performance across various metrics, including energy consumption, degree of imbalance, resource utilization, sensitivity analysis and task assignment computation time. Specifically, it significantly reduces computational time, achieving a reduction of 30–47% for fast Fourier transform applications and 44–48% for Gaussian elimination applications compared to existing algorithms. Additionally, it improves resource utilization, with enhancements ranging from 12.28 to 45.99% when compared to a variety of existing algorithms. To further validate the effectiveness of the proposed approach, it is applied to 10 real-world optimization problems. The performance is benchmarked against three state-of-the-art algorithms from the 'CEC2020 Competition': SASS, sCMAgES, and COLSHADE. This comprehensive evaluation framework provides a robust assessment of the algorithms' efficacy. [ABSTRACT FROM AUTHOR]
ISSN:14327643
DOI:10.1007/s00500-025-10614-y