Scholarlink

<p>The rapid growth of cloud services has led to a significant increase in energy consumption within large-scale data centers, underscoring the need for efficient resource management. Virtual machine (VM) placement plays a critical role in optimizing energy usage, balancing computational loads, and improving overall system performance. This paper introduces LAMOCS, a learning automata-based multi-objective cuckoo search algorithm designed to achieve optimal VM-to-server allocation in cloud environments. LAMOCS integrates reinforcement-driven learning automata with the multi-objective cuckoo search framework to enhance convergence efficiency and prevent premature stagnation in local optima. The proposed method simultaneously optimizes three key objectives: energy consumption, load balancing, and physical resource utilization. To validate performance, extensive simulations were conducted in MATLAB and compared against conventional genetic algorithm (GA) and particle swarm optimization (PSO) approaches. Results demonstrate that LAMOCS reduces energy consumption by approximately 7% and achieves 8% better load balancing on average, while also improving resource utilization. The key innovations of this work are the introduction of per-VM learning automata, tri-objective optimization across energy, load balancing, and utilization, and a balanced dual replacement strategy, which together distinguish LAMOCS from prior hybrid approaches. The outcomes confirm that LAMOCS provides an effective, scalable, and statistically robust solution for energy-aware VM placement in modern cloud data centers.</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

Reduction of energy consumption in cloud data centers with proper placement of virtual machines

  • 期刊论文

  • 发表日期：

    2025年11月23日

查看全文
  • Hamid Reza Naji,
  • Reza Esmaeili

摘要

The rapid growth of cloud services has led to a significant increase in energy consumption within large-scale data centers, underscoring the need for efficient resource management. Virtual machine (VM) placement plays a critical role in optimizing energy usage, balancing computational loads, and improving overall system performance. This paper introduces LAMOCS, a learning automata-based multi-objective cuckoo search algorithm designed to achieve optimal VM-to-server allocation in cloud environments. LAMOCS integrates reinforcement-driven learning automata with the multi-objective cuckoo search framework to enhance convergence efficiency and prevent premature stagnation in local optima. The proposed method simultaneously optimizes three key objectives: energy consumption, load balancing, and physical resource utilization. To validate performance, extensive simulations were conducted in MATLAB and compared against conventional genetic algorithm (GA) and particle swarm optimization (PSO) approaches. Results demonstrate that LAMOCS reduces energy consumption by approximately 7% and achieves 8% better load balancing on average, while also improving resource utilization. The key innovations of this work are the introduction of per-VM learning automata, tri-objective optimization across energy, load balancing, and utilization, and a balanced dual replacement strategy, which together distinguish LAMOCS from prior hybrid approaches. The outcomes confirm that LAMOCS provides an effective, scalable, and statistically robust solution for energy-aware VM placement in modern cloud data centers.