EUVIMEDEUVIMED entdecken
Crossref

Machine Learning-Based Intelligent Resource Allocation Framework for Dynamic Cloud Computing Environments

Gaurav Bhakuni and Dr. Sandeep Kumar

International Journal of Advanced Research in Science Communication and Technology · 19.06.2026 · 94

DOI: 10.48175/ijarsct-37113 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Advanced Research in Science Communication and Technology, online erschienen am 19.06.2026. Verfasst von Gaurav Bhakuni and Dr. Sandeep Kumar. Die Arbeit nennt 40 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Cloud computing has become the backbone of modern digital services, yet the highly dynamic and unpredictable nature of user workloads continues to make efficient resource allocation a persistent challenge. Conventional rule-based and heuristic provisioning strategies, such as round-robin, static thresholds, and metaheuristic optimizers, react to load changes only after they occur and therefore lead to either over-provisioning, which wastes energy and capital, or under-provisioning, which violates service-level agreements (SLAs). This paper proposes a Machine Learning-based Resource Allocation Framework (ML-RAF) that couples a Long Short-Term Memory (LSTM) workload forecaster with a Deep Reinforcement Learning (DRL) allocation agent to provision computing resources proactively in dynamic cloud environments. The LSTM module predicts short-horizon demand for CPU, memory, and bandwidth from historical monitoring traces, while the DRL agent learns an allocation policy that jointly optimizes resource utilization, response time, energy consumption, and SLA compliance. The framework was implemented and evaluated on a CloudSim-based simulation driven by a realistic workload trace, and benchmarked against four baselines: Round Robin, Threshold-based scaling, a Genetic Algorithm, and Particle Swarm Optimization. Experimental results show that the proposed ML-RAF improves average CPU utilization to 93%, reduces average response time by up to 67%, lowers data-center energy consumption by up to 37%, and decreases the SLA violation rate to 2.3%, clearly outperforming all baseline techniques. The findings confirm that combining predictive forecasting with learning-based decision making offers a practical and scalable path toward autonomous, energy-aware cloud resource management.

Zitieren

APA

Kumar, G. B. A. D. S. (2026). Machine Learning-Based Intelligent Resource Allocation Framework for Dynamic Cloud Computing Environments. International Journal of Advanced Research in Science Communication and Technology, 94. https://doi.org/10.48175/ijarsct-37113

Vancouver

Kumar GBADS. Machine Learning-Based Intelligent Resource Allocation Framework for Dynamic Cloud Computing Environments. International Journal of Advanced Research in Science Communication and Technology. 2026:94. doi:10.48175/ijarsct-37113

BibTeX
@article{Kumar2026,
  author = {Kumar, Gaurav Bhakuni and Dr. Sandeep},
  title = {Machine Learning-Based Intelligent Resource Allocation Framework for Dynamic Cloud Computing Environments},
  journal = {International Journal of Advanced Research in Science Communication and Technology},
  year = {2026},
  pages = {94},
  doi = {10.48175/ijarsct-37113}
}
Kostenlos starten →