Title: Extended: multi-objective resource optimisation using enhanced FPA-DRL in a heterogeneous cloud computing environment

Authors: Ramanpreet Kaur; Jagmeet Singh; Upinder Kaur; Karamjit Kaur; Amandeep Singh

Addresses: Department of Computer Application, Baba Farid College of Engineering and Technology, Bathinda, 151002, India ' Department of Computer Application, Baba Farid College of Engineering and Technology, Bathinda, 151002, India ' School of Computer Science and Engineering, Akal University, Talwandi Sabo, 151302, India ' Department of Computer Application, Baba Farid College of Engineering and Technology, Bathinda, 151002, India ' Department of Computer Application, Baba Farid College of Engineering and Technology, Bathinda, 151002, India

Abstract: Our proposed model is a multi-objective resource optimisation technique named MOROT used as MOROT enhanced flower pollination algorithm deep reinforcement learning (EFPA-DRL) for the purpose to handle the request having multi-objective. The EPFA algorithm, based on an extended DRL method, allows for the simultaneous optimisation of a large number of resources. The proposed model divided in 2 steps, first step is service analyser analyses the jobs, the cybershake seismogram prepares table of jobs based on size and time, minimum execution time (MET) create execution time based queue, In the second step jobs are received by EFPA-DRL for optimising local global region based on dynamic switching property. We used both synthetic and actual parallel workloads to evaluate the CloudSim toolset.

Keywords: resource allocation; optimisation; task scheduling; enhanced flower pollination; cloud technology.

DOI: 10.1504/IJAIGHR.2026.152914

International Journal of Artificial Intelligence Governance and Human Rights, 2026 Vol.1 No.1, pp.109 - 130

Received: 01 Jul 2025
Accepted: 04 Sep 2025
Published online: 14 Apr 2026 *