Scholarlink

<p>In cloud computing, scheduling workflows for data-intensive tasks is challenging due to task dependencies, heterogeneous resources, and high computational demands, all of which affect cost, execution time, and energy usage. This research proposes a novel hybrid optimization algorithm called Dolphin Swarm Sparrow Search Optimization (DSSSO) to address these challenges. The DSSSO model combines the broad search capability of Dolphin Swarm Optimization with the precise tuning ability of Sparrow Search Algorithm, enabling effective exploration and exploitation in the solution space. Unlike existing scheduling models that primarily optimize makespan and cost, the proposed DSSSO explicitly integrates energy consumption and resource utilization into a unified multi-objective framework. This dual-phase design—leveraging dolphins for global exploration and sparrows for local exploitation—ensures both convergence stability and adaptability in heterogeneous cloud environments. The main objective is to minimize makespan, reduce energy consumption, and lower scheduling cost while maximizing resource utilization. The model is evaluated using benchmark scientific workflows such as CyberShake, Montage, Epigenomics, and LIGO. Simulation results show that DSSSO outperforms existing methods, achieving up to 12% lower makespan, 15% less energy consumption, and 10% better resource usage compared to algorithms like Hybrid Bat Optimization and Improved Bat Optimization.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hybrid dolphin swarm sparrow search optimization based multi-objective cloud workflow scheduling

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月29日

查看全文
  • S. Subashree,
  • M. Rajakumaran,
  • G. Pushpa,
  • S. Senthilkumar

摘要

In cloud computing, scheduling workflows for data-intensive tasks is challenging due to task dependencies, heterogeneous resources, and high computational demands, all of which affect cost, execution time, and energy usage. This research proposes a novel hybrid optimization algorithm called Dolphin Swarm Sparrow Search Optimization (DSSSO) to address these challenges. The DSSSO model combines the broad search capability of Dolphin Swarm Optimization with the precise tuning ability of Sparrow Search Algorithm, enabling effective exploration and exploitation in the solution space. Unlike existing scheduling models that primarily optimize makespan and cost, the proposed DSSSO explicitly integrates energy consumption and resource utilization into a unified multi-objective framework. This dual-phase design—leveraging dolphins for global exploration and sparrows for local exploitation—ensures both convergence stability and adaptability in heterogeneous cloud environments. The main objective is to minimize makespan, reduce energy consumption, and lower scheduling cost while maximizing resource utilization. The model is evaluated using benchmark scientific workflows such as CyberShake, Montage, Epigenomics, and LIGO. Simulation results show that DSSSO outperforms existing methods, achieving up to 12% lower makespan, 15% less energy consumption, and 10% better resource usage compared to algorithms like Hybrid Bat Optimization and Improved Bat Optimization.