Year: 2026
Type: article
Abstract: The rapid expansion of cloud computing has significantly increased energy usage in data centers, leading to higher operational expenses and environmental consequences. This research introduces a hybrid solution that merges Ant Colony Optimization (ACO) and Deep Q-Learning (DQL) to enable energy-efficient and adaptable resource management in diverse multi-cloud settings. ACO generates energy-conscious task-to-resource mappings while DQL dynamically fine-tunes scheduling decisions in real-time, tackling challenges like workload variations, resource diversity and scalability issues. Simulation findings indicate that the combined ACO-DQL approach surpasses individual optimization methods and alternative scheduling techniques. The model achieved noteworthy outcomes, including an average task energy consumption of 0.33 J, CPU and memory utilization rates of 87% and 85%, respectively and an average task completion delay of 15 ms. These results validate that integrating optimization and reinforcement learning effectively reduces energy consumption, optimizes resource usage and ensures efficient, low-latency cloud operations. This proposed strategy offers a practical and scalable solution for eco-friendly cloud computing, meeting the rising demand for computationally intensive applications while lessening environmental impact.
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7204586526