Abstract Cloud-fog computing facilitates the implementation of massive scientific applications and IoT-driven applications by distributing the computation between heterogeneous cloud and edge resources. However, scheduling the workflow in such environments is still a tough, NP-hard problem, because of the heterogeneity of the resources, the dynamic workload, the latency constraints and the conflicting optimization objectives. This paper proposes a quantum-inspired multi-objective workflow scheduling framework based on a seahorse optimization strategy to efficiently allocate the tasks in distributed cloud-fog architectures. The workflow is modelled as Directed Acyclic Graph (DAG), and the scheduling problem is formulated to minimize the makespan and execution cost. The proposed model is implemented using workflowsim simulation environment and evaluated using five scientific workflows viz. Cybershake, Epigenomics, Montage, Inspiral, and Sipht. The experiments are performed over several different sizes of workflow and virtual machines. The results show that an average reduction of 9.28% in makespan and 11.49% in the execution cost when compared with BAT, PSO, SHO, GA-PSO and GWOA algorithms. The proposed approach also exhibits stable convergence behavior for large scale workflow instances. To overcome limitations of scalability of simulation-based evaluation, a machine learning based performance prediction module is added to make proactive scheduling decisions enabling capacity planning under large workload. The overall results show that the proposed framework offers robust, scalable, and cost-efficient workflow scheduling for heterogeneous cloud-fog environments.
Bansal et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: