ICOA: Enhanced Crayfish Optimization for Task Scheduling in Heterogeneous Cloud Environments

Abstract

Cloud computing systems are intended to efficiently process large amounts of data and provide ondemand computing resources. Scheduling tasks is one of the intrinsic issues of cloud computing, where tasks are allocated to resources and execution time is reduced while maximizing resource utilization. The present work proposes an improved version of the Crayfish Optimization Algorithm (COA), called ICOA, to solve the task scheduling problem in heterogeneous cloud computing. ICOA integrates chaotic mapping for population diversity, random opposition-based learning (ROBL) for enhanced global exploration, a non-linear control parameter for dynamic search balance, and a Cauchy mutation strategy to avoid premature convergence. Performance of the algorithm is evaluated using CloudSim with NASA, HPC2N, and a synthetic workload. Experimental comparisons on EHHO, MSA–CSA–PAES, ACO, and IWOA demonstrate that the proposed ICOA achieves remarkable improvements: makespan reduction of 50.8%, and resource utilization increase of 26.5%. These experimental results confirm that the proposed ICOA is an efficient and effective solution for task scheduling of complex cloud computing environments.