Reinforcement Learning for Efficient Resource Allocation in Cloud Computing: A Simulation Study
Keywords:
Cloud Computing, Resource Allocation, Reinforcement Learning, Deep Q-Network, Simulation, Cost Efficiency, Adaptive SchedulingAbstract
Cloud computing requires efficient resource allocation to ensure optimal utilization, performance, and cost reduction. Traditional methods are often insufficient to handle dynamic workloads. This paper presents a DRL approach using DQN to dynamically allocate cloud resources. A simulation environment models virtual machine provisioning and fluctuating workloads. The DQN agent learns resource allocation policies to maximize utilization and minimize operational cost and latency. Experimental results demonstrate improvements of 30% in CPU utilization, 40% reduction in latency, and 25% cost savings compared to static and threshold-based methods. The paper includes mathematical formulation, simulation data, and detailed analysis validating DRL’s effectiveness in cloud resource management.
📊 Article Downloads
References
Abadi, M., et al. (2016). TensorFlow: Large-scale machine learning on heterogeneous systems.
Armbrust, M., et al. (2010). A view of cloud computing. Communications of the ACM, 53(4), 50-58.
Bai, X., et al. (2017). Dynamic resource allocation for cloud computing using reinforcement learning. Future Generation Computer Systems, 72, 419-429.
Chen, X., et al. (2019). Efficient multi-user computation offloading. IEEE/ACM Transactions on Networking, 24(5), 2795-2808.
Jin, H., et al. (2017). Cloud resource provisioning based on reinforcement learning. IEEE ICWS, 243-250.
Kumar, R., et al. (2016). Intelligent resource provisioning in cloud computing. Int J Cloud Computing & Services Science, 5(2), 132-140.
Mnih, V., et al. (2015). Human-level control through deep reinforcement learning. Nature, 518(7540), 529–533.
Xu, J., et al. (2014). Fuzzy modeling in virtualized data center management. IEEE Cloud, 296-303.
Zhang, Q., et al. (2020). Cloud computing: state-of-the-art and research challenges. Journal of Internet Services and Applications, 1(1), 7-18.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 International Journal of Scientific Research in Science and Technology
This work is licensed under a Creative Commons Attribution 4.0 International License.
https://creativecommons.org/licenses/by/4.0
