Energy efficient resource allocation using deep reinforcement learning with starling murmuration optimization in cloud

Abstract

Cloud computing is one of the widely used technologies due to its advanced features like pay -per-use, scalability, flexibility, and many more. It initially aims at fulfilling users’ needs for accessing computing resources and purchasing cloud services as necessary. This is all done within the framework of on-demand resource sharing, which is facilitated by internet-based applications. Efficient resource allocation (RA) in the cloud includes challenges of high-dimensional state spaces and scalability, as the tasks keep increasing. To solve this problem, deep reinforcement learning (DRL) with Taylor-based optimization neighborhood starling murmuration optimization (TBONSMO) algorithm model is used for RA according to the tasks efficiently. The integration of DRL with deep neural networks enhances task scheduling by enabling adaptive and efficient RA. This approach improves the decision-making processes, optimizes energy usage as well as system performance in response to varying workloads and operational demands. The experimental results of the proposed TBONSMO method were calculated as follows: energy consumption of 30.58 kilo watt per hour and resource utilization of 99.54%, which is lesser when compared to the existing deep Q-network (DQN) approach.