PTBO: Energy and Makespan-Aware Scheduling in Cloud Computing Using Meta-Heuristic Algorithm

Efficient task scheduling in cloud computing environments is essential to optimize resource utilization, minimize execution time, and reduce operational costs. Traditional scheduling algorithms often struggle with balancing multi-objective constraints such as makespan, energy consumption, and load distribution, especially under dynamic workloads. To address these challenges, this paper presents a novel Painting Training Based Optimization (PTBO) algorithm for efficient task scheduling in cloud computing, aiming to minimize makespan and energy consumption while enhancing resource utilization. Inspired by the structured and iterative learning process of artists, PTBO integrates exploration and exploitation strategies to achieve optimal task-to-VM mapping. The algorithm is evaluated using the HPC2N workload dataset in a CloudSim simulation environment and compared against widely used metaheuristic algorithms GA, PSO, and ACO. Experimental results reveal that PTBO reduces makespan by 24.2 % and energy consumption by 29.6 % on average across varying task volumes. These improvements highlight PTBO's capability to outperform existing techniques in dynamic cloud environments, making it a promising solution for scalable, energy-efficient, and time-sensitive task scheduling.