Scholarlink

Cloud computing stands as the dominant model which provides adjustable computing resources that automatically respond to user needs. Efficient resource allocation in cloud systems needs sophisticated algorithms to manage dynamic workloads with reduced power usage while achieving maximum resource utilization. This research work examines the performance of nature-inspired metaheuristic algorithms for Infrastructure-as-a-Service (IaaS) cloud resource allocation using the Energy Valley Optimizer (EVO). The study conducts its experiments through CloudSim which provides a powerful cloud simulation system to develop practical cloud computing systems. Two experimental settings consist of 5 virtual machines (VMs) and 10 VMs under varying workload sizes from 100 to 500 tasks to 1000–5000 tasks. Performance assessment includes execution duration together with power consumption measurements in kilowatt-hours (KWh) units and resource utilization percentage. The obtained results demonstrate that EVO provide superior performance through EVO’s 15% shorter execution time and 20–25% lower power usage than Symbiotic Organisms Search (SOS) and Particle Swarm Optimization (PSO) and Cuckoo Search (CSA).

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

Scalable Resource Allocation for Cloud IaaS Using Energy Valley Algorithm

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Punit Gupta,
  • Dhruvil Jariwala,
  • Shubham Joshi,
  • Mayank Kumar Goyal

摘要

Cloud computing stands as the dominant model which provides adjustable computing resources that automatically respond to user needs. Efficient resource allocation in cloud systems needs sophisticated algorithms to manage dynamic workloads with reduced power usage while achieving maximum resource utilization. This research work examines the performance of nature-inspired metaheuristic algorithms for Infrastructure-as-a-Service (IaaS) cloud resource allocation using the Energy Valley Optimizer (EVO). The study conducts its experiments through CloudSim which provides a powerful cloud simulation system to develop practical cloud computing systems. Two experimental settings consist of 5 virtual machines (VMs) and 10 VMs under varying workload sizes from 100 to 500 tasks to 1000–5000 tasks. Performance assessment includes execution duration together with power consumption measurements in kilowatt-hours (KWh) units and resource utilization percentage. The obtained results demonstrate that EVO provide superior performance through EVO’s 15% shorter execution time and 20–25% lower power usage than Symbiotic Organisms Search (SOS) and Particle Swarm Optimization (PSO) and Cuckoo Search (CSA).