Deep Reinforcement Learning for Cloud Resource Provisioning

Authors: Perumallapalli Ravikumar

Country: USA

Full-text Research PDF File:   View   |   Download

Abstract: Efficient resource provisioning in cloud computing environments is critical to optimizing cost, performance, and energy efficiency. Traditional resource allocation techniques often struggle to adapt to the dynamic and complex nature of cloud workloads, leading to suboptimal utilization. Deep Reinforcement Learning (DRL) has emerged as a promising approach for tackling this challenge, leveraging its ability to learn optimal policies through trial-and-error interactions with the environment. This paper explores the application of DRL to cloud resource provisioning, presenting a framework that dynamically allocates resources based on workload demands.

By modeling the cloud environment as a Markov Decision Process (MDP), DRL agents learn policies to minimize provisioning costs while ensuring Quality of Service (QoS) requirements. Simulation results demonstrate that the proposed approach outperforms traditional heuristics, achieving higher resource utilization and cost savings in diverse workload scenarios. The study highlights the potential of DRL as a scalable and adaptive solution for managing cloud resources in increasingly complex and dynamic cloud infrastructures.

Keywords: -

Paper Id: 231800

Published On: 2016-01-05

Published In: Volume 4, Issue 1, January-February 2016

Cite This: Deep Reinforcement Learning for Cloud Resource Provisioning - Perumallapalli Ravikumar - IJIRMPS Volume 4, Issue 1, January-February 2016.

Share this