GWO-Based Workflow Scheduling in Cloud-Fog Environments

The scientific workflows are a set of predefined tasks aiming to achieve a target goal state. Cloud computing provides an appropriate environment for orchestrating such scientific workflows. However, scheduling these workflows in a cloud environment is a big challenge due to the high latency and connectivity on low bandwidth network infrastructure. Nowadays, fog and edge computing devices offer an alternative to cloud computing. As a result, these devices can provide cloud computing capabilities for processing scientific workflows near the source. This paper explores the possibilities of scheduling workflows on both cloud and fog computing devices. In order to schedule workflows on cloud-fog integrated devices, we implemented the Grey Wolf Optimizer (GWO) based algorithm using dynamic weights for updating the grey wolf position on the FogWorkflowSim simulator. The proposed scheme avoids getting into a local optimum and improves the performance of the QoS parameters. We also compare the outcomes of different scheduling algorithms in cloud and hybrid cloud-fog architectures. The simulation results of the proposed approach using well-known scientific workflows suggest that cloud-fog integration for workflow scheduling improves processing performance while reducing energy consumption and cost.