Scholarlink

<p>Effective process scheduling is mandatory in cloud computing. It optimizes the use of resources, reduces the execution time, and reduces the operating cost. This study introduces a new workflow scheduling algorithm based on the salp swarm algorithm (SSA). This algorithm is a nature-based metaheuristic and is based on salp swimming behavior in the ocean. The proposed scheduler is based on the use of a single-objective fitness function to optimize the overall workflow makespan. It solves the NP-hard cloud scheduling problem, which involves effective mapping of tasks to heterogeneous virtual machines. The workflow execution is represented via a directed acyclic graph (DAG). The SSA is very efficient in search space exploration and is insensitive to parameter choices. To evaluate the quality of the proposed scheduler, several experiments were conducted via the WorkflowSim simulation framework. The makespan and energy consumption are taken as important performance indicators in the assessment. The WS-SSA was evaluated against FCFS, MCT, MIN-MIN, MAX-MIN, round robin, the WOA, the GA, and PSO under identical VM settings and workflow inputs to ensure fairness. It reduces the average makespan by up to 38% compared with classical methods and approximately 8.7% compared with the WOA and approximately 5–22% and 3–12% over the GA and PSO, respectively.</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

WS-SSA: workflow scheduling in cloud computing using salp swarm algorithm

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月24日

查看全文
  • Aya A. Sharawy,
  • Rasha H. Sakr,
  • Waleed Eladrosy,
  • Mohammed F. Alrahmawy

摘要

Effective process scheduling is mandatory in cloud computing. It optimizes the use of resources, reduces the execution time, and reduces the operating cost. This study introduces a new workflow scheduling algorithm based on the salp swarm algorithm (SSA). This algorithm is a nature-based metaheuristic and is based on salp swimming behavior in the ocean. The proposed scheduler is based on the use of a single-objective fitness function to optimize the overall workflow makespan. It solves the NP-hard cloud scheduling problem, which involves effective mapping of tasks to heterogeneous virtual machines. The workflow execution is represented via a directed acyclic graph (DAG). The SSA is very efficient in search space exploration and is insensitive to parameter choices. To evaluate the quality of the proposed scheduler, several experiments were conducted via the WorkflowSim simulation framework. The makespan and energy consumption are taken as important performance indicators in the assessment. The WS-SSA was evaluated against FCFS, MCT, MIN-MIN, MAX-MIN, round robin, the WOA, the GA, and PSO under identical VM settings and workflow inputs to ensure fairness. It reduces the average makespan by up to 38% compared with classical methods and approximately 8.7% compared with the WOA and approximately 5–22% and 3–12% over the GA and PSO, respectively.