The Application of Hybrid Krill Herd Artificial Hummingbird Algorithm for Scientific Workflow Scheduling in Fog Computing

Abstract

Fog Computing (FC) provides processing and storage resources at the edge of the Internet of Things (IoT). By doing so, FC can help reduce latency and improve reliability of IoT networks. The energy consumption of servers and computing resources is one of the factors that directly affect conservation costs in fog environments. Energy consumption can be reduced by efficacious scheduling methods so that tasks are offloaded on the best possible resources. To deal with this problem, a binary model based on the combination of the Krill Herd Algorithm (KHA) and the Artificial Hummingbird Algorithm (AHA) is introduced as Binary KHA- AHA (BAHA-KHA). KHA is used to improve AHA. Also, the BAHA-KHA local optimal problem for task scheduling in FC environments is solved using the dynamic voltage and frequency scaling (DVFS) method. The Heterogeneous Earliest Finish Time (HEFT) method is used to discover the order of task flow execution. The goal of the BAHA-KHA model is to minimize the number of resources, the communication between dependent tasks, and reduce energy consumption. In this paper, the FC environment is considered to address the workflow scheduling issue to reduce energy consumption and minimize makespan on fog resources. The results were tested on five different workflows (Montage, CyberShake, LIGO, SIPHT, and Epigenomics). The evaluations show that the BAHA-KHA model has the best performance in comparison with the AHA, KHA, PSO and GA algorithms. The BAHA-KHA model has reduced the makespan rate by about 18% and the energy consumption by about 24% in comparison with GA.