Scheduling Linear Workflows with Multicriteria based Partial Computations in a Cloud–Fog–Mist Environment

In distributed computing architectures, workloads often comprise jobs with varying levels of security requirements. In such environments, it is crucial to implement security-aware scheduling schemes to ensure the appropriate execution of applications. This research focuses on security-aware scheduling techniques for linear workflow (LW) applications within a cloud-fog-mist computing platform. Mist resources are allocated to process LW jobs with the highest security requirements, classified as mist jobs. LW jobs identified as fog jobs possess intermediate security levels and can execute on either fog or mist resources. LW jobs without security requirements can run on cloud or fog resources; however, when executed on fog resources, they may undergo partial computations based on multiple criteria to prioritize fog jobs and ensure timeliness. Extensive simulation experiments were conducted to evaluate the impact of partial computations on performance. The results provide valuable insights into how dynamic adjustments of partial computations during runtime affect system performance across different scenarios.