Scholarlink

<p>Energy-efficient virtual machine (VM) migration could be a crucial challenge towards optimizing resource overusage for energy consumption. This paper discusses an efficient method for VM migration based on Q-learning toward dynamic workload balancing of resources across Physical Machines. The research includes energy-aware decision-making and adaptive threshold incorporation for SLA fulfilment and usage minimization. Experimentation results prove the efficiency of the current approach for state-of-the-art algorithms like Updated SESA, Enhanced Dragonfly, and Optimal Meta-Heuristic Elastic Scheduling (OMES). The implementation reduces power consumption by up to 16.53% when compared with the OMES approach, along with a service level agreement violation improvement of about 50% for 500 individual VMs. These results show the strength and efficiency of the proposed work and also emphasize the application of reinforcement learning for addressing critical challenges in cloud computing. It further provides future directions in energy-efficient resource management.</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

Energy-efficient VM selection using deep neural networks for scalable cloud computing

  • 期刊论文

  • 发表日期：

    2026年7月24日

查看全文
  • Harpreet Kaur,
  • Jaswinder Singh,
  • Harmandeep Singh

摘要

Energy-efficient virtual machine (VM) migration could be a crucial challenge towards optimizing resource overusage for energy consumption. This paper discusses an efficient method for VM migration based on Q-learning toward dynamic workload balancing of resources across Physical Machines. The research includes energy-aware decision-making and adaptive threshold incorporation for SLA fulfilment and usage minimization. Experimentation results prove the efficiency of the current approach for state-of-the-art algorithms like Updated SESA, Enhanced Dragonfly, and Optimal Meta-Heuristic Elastic Scheduling (OMES). The implementation reduces power consumption by up to 16.53% when compared with the OMES approach, along with a service level agreement violation improvement of about 50% for 500 individual VMs. These results show the strength and efficiency of the proposed work and also emphasize the application of reinforcement learning for addressing critical challenges in cloud computing. It further provides future directions in energy-efficient resource management.