PKO: Energy and Makespan Aware Scheduling in Cloud Environment using Metaheuristic Algorithm
Santhosh Kumar Medishetti, Rameshwaraiah Kurupati, Golamari Sravan Kumar, Nalavath Sai Charan, Minkikar Sai Krishna, M. S. Subodh Raj
Abstract
This study introduces a novel task scheduling algorithm for cloud computing based on the Pied Kingfisher Optimizer (PKO), inspired by the bird’s precision and adaptability in dynamic environments. Implemented on Google Cloud workloads and simulated using CloudSim, PKO is designed to optimize three key performance metrics: makespan, energy consumption, and throughput. The algorithm's effectiveness is evaluated against three widely used metaheuristic algorithms such as GA, PSO, and IBOA. Experimental results demonstrate that PKO consistently achieves greater improvement of makespan and energy consumption by 24.91% and 33.16% while maximizing throughput by 31.18% across diverse workload scenarios. Its hybrid exploration–exploitation strategy enables better task-to-resource mapping, dynamically adjusting to workload fluctuations and infrastructure constraints. Compared to GA and PSO, PKO shows improved convergence speed and scheduling accuracy, while surpassing IBOA in energy-aware throughput optimization. The integration of PKO in the cloud environment reveals its scalability and robustness for real-time scheduling challenges. This makes it a promising approach for efficient, sustainable, and performance-aware task scheduling in cloud infrastructures, particularly under heterogeneous and time-sensitive workloads.