Document Type : Research Article
Authors
1
Department of Computer Science, Shahid Bahonar University of Kerman, Kerman, Iran
2
Shahid Bahonar University of Kerman
3
Department of Computer Science ,Shahid Bahonar University,Kerman,Iran
4
Computer Science Department, Shahid Bahonar University of Kerman, Kerman, Iran
10.22060/miscj.2025.23520.5381
Abstract
During scheduling process, it is important to respect the constraints given by the jobs and the cloud providers. In addition to maintaining a balance between Quality of Service (QoS), fairness, and efficiency of jobs, scheduling is challenging. This paper aims to propose an efficient algorithm for load-balanced task scheduling in the cloud. Our algorithm uses a new meta-heuristic algorithm called COA (Coati Optimization Algorithm) to solve the task scheduling problem. This method is called COTSA (Coati Optimization-based Task Scheduling Algorithm). Its main goal is to reduce execution costs, load balancing, resource consumption, and makepan. Additionally, experimental results indicate that COTSA contributes to reduced energy consumption and enhanced system scalability and fault tolerance under simulated conditions. These improvements suggest potential suitability for dynamic and large-scale cloud infrastructures, though performance may vary depending on workload characteristics and system configurations. It is compared with Walrus Optimizer (WO), Slap Swarm Algorithm (SSA), Whale Optimization Algorithm (WOA), Zebra Optimization Algorithms (ZOA), Grasshopper Optimization Algorithm (GOA), Sooty Tern Optimization Algorithm (STOA), Golden Eagle Optimizer (GEO), Grey Wolf Optimizer (GWO), Subtraction-Average-Based Optimizer (SABO), and Sand Cat Swarm Optimization (SCSO), which are popular meta-heuristics. Experimental results demonstrate that COTSA reduces makespan by approximately 9%, lowers execution cost by up to 40%, improves resource utilization by around 3%, and enhances load balance by up to 30%, energy consumption about 36%, scalability near 17%, and fault tolerance about 16%, making it a robust and scalable solution for efficient cloud task scheduling.
Keywords
Main Subjects
)