Energy-Efficient Resource Allocation in Cloud Computing Environments Using Machine Learning

Abstract

  

Cloud computing has become a fundamental technology for delivering scalable and on-demand computing resources to diverse applications. However, the rapid growth of cloud data centers has significantly increased energy consumption, operational costs, and environmental impact. Efficient resource allocation is therefore essential to improve server utilization while maintaining Quality of Service (QoS). Traditional scheduling algorithms often rely on static or heuristic approaches that are unable to adapt effectively to dynamic workload variations, resulting in resource underutilization and unnecessary power consumption. This paper proposes an Energy-Efficient Machine Learning-Based Resource Allocation (EEMLRA) framework that employs supervised machine learning to predict workload demand and perform proactive virtual machine allocation. The proposed framework integrates workload monitoring, feature extraction, prediction, and dynamic resource allocation to optimize energy efficiency while reducing Service Level Agreement (SLA) violations. Performance evaluation is conducted using CloudSim Plus with representative cloud workload traces. Simulation results indicate that the proposed framework reduces energy consumption, improves CPU utilization, minimizes VM migrations, and enhances response time compared with conventional scheduling approaches. The proposed method offers a scalable and sustainable solution for intelligent cloud resource management.