OSSA Scheduler: Opposition-Based Learning Salp Swarm Algorithm for Task Scheduling in Cloud Computing

Abstract

In cloud computing, task scheduling has a direct influence on service quality. Task scheduling means allocating tasks to available resources based on user specifications. This NP-hard problem seeks to develop an optimal scheduler for resource allocation to complete tasks in the shortest amount of time achievable. Several methods have been presented to tackle the task scheduling issue. In this study, an Opposition-based Learning Salp Swarm Algorithm (OSSA) to address task scheduling issues. The initial population phase of the proposed OSSA scheduler for task scheduling in a cloud computing environment uses Opposition-Based Learning (OBL) to minimize execution time. OBL generates a diversified and high-quality initial population, improving the optimization process's overall performance. The paper compares the proposed OSSA algorithm to various metaheuristic algorithms, like the standard Salp Swarm Algorithm (SSA), Differential Evolution (DE) and Sine Cosine Algorithm (SCA). The results shows that the OSSA algorithm can solve the task scheduling problem more efficiently and achieve superior solutions for minimizing the makespan.