Joint Optimization of Scheduling Length and Cost Based on White Shark Optimization in Heterogeneous Clouds

In the era of the Internet of Things, the significant increase in data volume, time, and space complexity presents great challenges to workflow scheduling in resource-constrained clouds. This study proposes an efficient hybrid algorithm, denoted white shark optimization (WSO) algorithm with budget constraints (BC-WSO), designed to adhere to budget constraints. The primary objective of BC-WSO is to optimize the scheduling length and cost. This objective is achieved by employing a heuristic algorithm that utilizes the predicted makespan matrix (PMMS) alongside the WSO algorithm as its foundation. The PMMS can minimize the scheduling length of a workflow application and satisfy the task prioritization dependencies. BC-WSO incorporates PMMS into the population initialization phase to improve the accuracy of WSO and accelerate the convergence process. Extensive experiments in two real-world scientific workflow applications show that BC-WSO outperforms current state-of-the-art meta-heuristic algorithms in simultaneously optimizing scheduling length and cost.