Heuristic Scheduling Algorithm for Workflow Applications in Cloud-Fog Computing Based on Realistic Client Port Communication

Cloud-fog computing systems require task scheduling to efficiently manage resources and ensure timely task execution in a distributed environment. Fog computing extends computation, storage, and networking to the edge of the network, thereby improving the user experience, and reducing latency, and data transfer. Task scheduling and client-processing resource management in cloud-fog computing affect cloud system performance and client services. Currently, task scheduling methods mainly focus on assigning client tasks to the most suitable cloud resource or virtual machine (VM) to obtain the minimum scheduled time. However, these techniques do not consider a realistic communication model that, allows many parallel tasks to be sent and received simultaneously without delay. This paper introduces a heuristic workflow scheduling algorithm called Collision-Avoid based on Cloud task Scheduling (CACS) to assign client workflow represented by a directed acyclic graph (DAG) and execute the most appropriate cloud-fog nodes under the single client I/O port restriction to achieve minimum workflow completion time, least delay, and higher resource performance. FogWorkflowSim was used as the cloud-fog computing system simulation. The experimental results showed that the proposed method achieved much better scheduling results than the other compared algorithms.