Hybrid Heuristic and Particle Swarm Optimization Approach to Cloud Task Scheduling

Scheduling tasks on cloud systems is a critical optimization problem that aims to distribute tasks among available resources in the most effective way. This issue falls under the category of NP-hard problems and generating exact and deterministic solutions requires high computational costs. Metaheuristic approaches have proven to provide successful results in solving such problems. Particle Swarm Optimization (PSO), one of these algorithms, is a widely used method in the literature due to its advantages, such as fast convergence, simple applicability, and low computational cost. In this study, a hybrid heuristic-based Particle Swarm Optimization approach is proposed to improve task scheduling efficiency. The proposed approach improves the solution quality by integrating a heuristic mechanism into the random population generation process of PSO. In comparison to First Come First Serve, Ant Colony Optimization (ACO), and conventional PSO, the suggested approach delivers better makespan and reduced energy consumption, according to the simulation analysis carried out in the CloudSim simulation environment. According to simulations, Heuristic PSO outperforms traditional PSO and ACO methods in terms of makespan time, reducing it by an average of 61.42% and 62.84%, respectively. It also uses 26.18% less energy than PSO and 27.33% less than ACO, according to its energy consumption data. The results show that the suggested method offers a more effective substitute for scheduling tasks in cloud computing systems.