Hybrid Cloud Workflow Scheduling Algorithm Based on Ant Colony Optimization

The rapid development of digital technologies has made digital transformation a core trend across various industries. Cloud computing provides flexible computing power to support digital transformation, however, single cloud architecture struggles to balance efficiency and privacy security. As a result, hybrid cloud has emerged as the mainstream cloud solution for data-intensive scenarios. By combining the scalability and cost-effectiveness of public clouds with the security and privacy of private clouds, hybrid cloud presents significant application potential and practical value in large-scale scientific workflow domains. However, scheduling privacy-sensitive tasks on public cloud platforms is not always feasible, and cross-cloud data transmission in hybrid cloud environments poses risks of privacy data leakage. To address these challenges, this paper proposes a hybrid cloud workflow scheduling algorithm based on the Ant Colony Optimization algorithm (HACO), which considers privacy-sensitive tasks. Under deadline and budget constraints in a hybrid cloud environment, the proposed algorithm establishes a data security model that accounts for the security of user data transmission. By leveraging the upward rank of tasks, it dynamically allocates budgets to tasks, thereby minimizing the overall workflow execution cost while satisfying deadline and budget constraints. Additionally, this paper designs an efficient virtual machine (VM) selection strategy that selects appropriate VMs based on the data security requirements of each workflow task while considering cost and makespan, proposed approach effectively reduces execution costs. Finally, using four real-world scientific workflows and randomly generated workflows, the proposed HACO algorithm is compared with PSSA and ACO. Experimental results demonstrate that HACO achieves the optimal overall task execution cost while ensuring workflow data security.