Scholarlink

<p>Cloud resource management is challenged by the quick development of technology and customer demand, especially when it comes to optimizing virtual machine (VM) migration and usage. In order to provide real-time responsiveness without relying on pre-existing structures, this paper presents an Alterable Clustering Scheme (ACS) that leverages virtual machine migration to adjust to high-demand scenarios. ACS uses a linear training strategy to reduce allocation failures and deep learning techniques to assess clustering and migration adaptability during periods of peak demand. An extended security–performance evaluation contrasting the encryption techniques of RSA, AES, Blowfish, and 3DES is carried out to guarantee safe and effective operation. According to the results, AES and Blowfish perform better than RSA and 3DES in terms of latency, throughput, and encryption speed, thus making them more suitable for dynamic cloud resource management processes such as allocation, adaptation, and failure analysis. Additionally, the experimental evaluation has been expanded to large-scale environments using CloudSim 7G and OpenStack testbeds, incorporating recent ML and metaheuristic baselines. Results show that ACS achieves up to 15–25% better resource utilization, 20% lower migration costs, and 10% fewer SLA violations compared to state-of-the-art methods. Experimental results demonstrate that ACS enhances resource allocation efficiency by increasing allocation rates by 9.29%, reducing migration ratios by 15.63%, decreasing allocation failures by 9.78%, and shortening allocation times by 13.37%. Overall, ACS demonstrates scalability, adaptability, and security robustness, offering a practical framework for intelligent cloud 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

Enhancing cloud resource management with dynamic clustering scheme: a deep learning approach for optimized VM migration and secured data protection

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月17日

查看全文
  • Jeny Jijo,
  • Saritha,
  • Bhavya Kallapu,
  • N. S. Krishnaraj Rao,
  • Rama Moorthy Hejamadi,
  • Raghunandan Kemmannu Ramesh,
  • Lubna A. Gabralla

摘要

Cloud resource management is challenged by the quick development of technology and customer demand, especially when it comes to optimizing virtual machine (VM) migration and usage. In order to provide real-time responsiveness without relying on pre-existing structures, this paper presents an Alterable Clustering Scheme (ACS) that leverages virtual machine migration to adjust to high-demand scenarios. ACS uses a linear training strategy to reduce allocation failures and deep learning techniques to assess clustering and migration adaptability during periods of peak demand. An extended security–performance evaluation contrasting the encryption techniques of RSA, AES, Blowfish, and 3DES is carried out to guarantee safe and effective operation. According to the results, AES and Blowfish perform better than RSA and 3DES in terms of latency, throughput, and encryption speed, thus making them more suitable for dynamic cloud resource management processes such as allocation, adaptation, and failure analysis. Additionally, the experimental evaluation has been expanded to large-scale environments using CloudSim 7G and OpenStack testbeds, incorporating recent ML and metaheuristic baselines. Results show that ACS achieves up to 15–25% better resource utilization, 20% lower migration costs, and 10% fewer SLA violations compared to state-of-the-art methods. Experimental results demonstrate that ACS enhances resource allocation efficiency by increasing allocation rates by 9.29%, reducing migration ratios by 15.63%, decreasing allocation failures by 9.78%, and shortening allocation times by 13.37%. Overall, ACS demonstrates scalability, adaptability, and security robustness, offering a practical framework for intelligent cloud resource management.