Networked Sensing Systems

5Energy-Aware System for Dynamic Workflow Scheduling in Cloud Data Centers: A Genetic Algorithm with DQN Approach

Hariharan B.¹*, Anupama C.G.¹, Ratna Kumari Neerukonda¹ and Rajesh Kumar Dhanaraj²

¹Department of Computational Intelligence, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India

²Symbiosis Institute of Computer Studies and Research (SICSR), Symbiosis International (Deemed University), Pune, India

Abstract

Due to the rising services offered by data centers, such as load balancing, auto-scaling, dynamic resource allocation, and efficient resource use, many firms are migrating their businesses to the cloud. Servers are considered the essential processing units in a data center. The growing expansion of cloud applications is leading to a significant rise in energy usage and increased carbon gas emissions. Upon receiving a user’s request, the scheduler allocates the tasks to cloud resources for execution. Efficiently organizing tasks in a cloud environment poses a formidable challenge. Several sophisticated job schedulers exist for managing workflow; their architecture mostly caters to batch workflow rather than real-time workflow. Efficient allocation of real-time workflows to suitable virtual machines on a server requires implementing a dynamic workflow scheduling technique that effectively minimizes energy usage. This study proposes an alternative approach to dynamic workflow scheduling, aiming to improve the workflow schedule to decrease both ...