Analyze and Optimize Cloud Resource Allocation Using Machine Learning for Sustainable Green IT Infrastructure

Abstract

The rapid growth of cloud computing has resulted in significant energy consumption, raising sustainability concerns and environmental impacts associated with large-scale IT infrastructure. Efficient cloud resource allocation is essential for balancing performance, cost, and energy efficiency while supporting dynamic workloads. This study investigates the use of machine learning algorithms to optimize cloud resource allocation in multi-tenant cloud environments to promote sustainable green IT infrastructure. A conceptual model was developed linking predictive workload modeling, intelligent VM placement, energy-aware scheduling, and dynamic scaling with key sustainability outcomes, including energy consumption reduction, carbon footprint minimization, and resource utilization efficiency. A quantitative research design employing Partial Least Squares Structural Equation Modeling was applied to assess relationships between these constructs. Data were collected from 420 cloud engineers, IT infrastructure managers, and data center operators across enterprises implementing green IT initiatives. The measurement model demonstrated reliability and convergent validity with composite reliability values exceeding 0.91 and average variance extracted above 0.63. Structural model analysis revealed that predictive workload modeling beta 0.56 p < 0.001, energy-aware scheduling beta 0.49 p < 0.001, and intelligent VM placement beta 0.42 p < 0.001 significantly enhance energy efficiency and resource utilization. Dynamic scaling mediates the relationship between workload prediction and energy efficiency beta 0.44 p < 0.001. The model explained 65 percent of variance in energy efficiency and 61 percent in sustainable performance. These findings indicate that integrating machine learning for workload forecasting, intelligent allocation, and energy-aware scheduling significantly supports sustainable cloud operations. The study provides a validated framework for cloud infrastructure managers and policymakers to design energy-efficient, scalable, and environmentally sustainable cloud computing environments.