An Efficient Task Scheduling Approach in Cloud Computing Using Hybrid Fruit Fly and Ant Colony Optimization Techniques

Abstract

Although cloud computing offers on-demand resources, effective job scheduling is still a major challenge to increase efficiency and resource usage. For large-scale, dynamic workloads, traditional algorithms s FCFS and Min-Min are straightforward yet ineffective. In order to generate high-quality task–VM mappings, this study suggests a hybrid nature-inspired algorithm called fruit fly optimization–ant colony optimization (FOA-ACO), which combines the exploitative ant colony optimization (ACO) and the exploratory fruit fly optimization algorithm (FOA). Minimizing makespan, optimizing resource usage, and distributing load evenly among virtual machines (VM) are the objectives. Workload traces from PlanetLab and CloudSim 3.0.3 are used to assess the method in a variety of experimental scenarios involving up to 1000 jobs and numerous diverse VM. The suggested FOA-ACO methodology enhances overall cloud performance by decreasing the makespan by 9.57%, employing more resources by 7.14%, and enhancing the load balancing factor (LBF) by 23.06%. This was observed by comparing it with other approaches such as FCFS, Min-Min, solo ACO, and WOA. An effective solution to the scheduling problems in today’s cloud environments is provided by this hybrid method. Future research may concentrate on refining the system to take into consideration various objectives, such as cost and reliability, enhancing its use of energy, and optimizing its performance with very massive cloud setups. Also, the approach might be refined to facilitate decision-making in real time and incorporate machine learning methods for greater flexibility.