PSO-Based Workflow Scheduling: a Comparative Evaluation of Cloud and Cloud-Fog Environments

EasyChair Preprint 6968

Abstract

This paper presents a comparative evaluation of cloud and cloud-fog environments for workflow scheduling using the classical Particle Swarm Optimization (PSO) algorithm. This paper also presents a weighted sum objective function based on three objectives: makespan, cost and energy. The recently proposed FogWorkflowSim is used as the simulation environment for the cloud and cloud-fog architectures. The results obtained for two well-known scientific workflows (Montage and CyberShake) show that the incorporation of the fog layer for the execution of workflows has the potential to improve processing efficiency and reduce energy consumption, motivating the cloud-fog computing paradigm. Future work will focus on the evaluation of other types of workflows such as Epigenomics, LIGO and SIPHT as well as increasing the number of tasks in a workflow.

Keyphrases: Cloud Computing, Fog Computing, Particle Swarm Optimization, workflow scheduling

@booklet{EasyChair:6968,
  author    = {Dineshan Subramoney and Clement Nyirenda},
  title     = {PSO-Based Workflow Scheduling: a Comparative Evaluation of Cloud and Cloud-Fog Environments},
  howpublished = {EasyChair Preprint 6968},
  year      = {EasyChair, 2021}}