A Novel Hybrid Algorithm Based on Butterfly and Flower Pollination Algorithms for Scheduling Independent Tasks on Cloud Computing

Abstract

Cloud computing is an Internet-based computing paradigm where virtual servers or workstations are offered as platforms, software, infrastructure, and resources. Task scheduling is considered one of the major NP-hard problems in cloud environments, posing several challenges to efficient resource allocation. Many metaheuristic algorithms have been extensively employed to address these task-scheduling problems as discrete optimization problems and have given rise to some proposals. However, these algorithms have inherent limitations due to local optima and convergence to poor results. This paper suggests a hybrid strategy for organizing independent tasks in heterogeneous cloud resources by incorporating the Butterfly Optimization Algorithm (BOA) and Flower Pollination Algorithm (FPA). Although BOA suffers from local optima and loss of diversity, which may cause an early convergence of the swarm, our hybrid approach outperforms such weaknesses by exploiting a mutualism-based mechanism. Indeed, the proposed hybrid algorithm outperforms existing methods while considering different task quantities with better scalability. Experiments are conducted within the CloudSim simulation framework with many task instances. Statistical analysis is performed to test the significance of the obtained results, which confirms that the suggested algorithm is effective at solving cloud-based task scheduling issues. The study findings indicate that the hybrid metaheuristic algorithm could be a promising approach to improving resource utilization and optimizing cloud task scheduling.