A novel workflow scheduling with multi-criteria using particle swarm optimization for heterogeneous computing systems

Importance of workflow applications (WAs) is expediting in various fields of science and engineering. Scheduling of WAs is a non-deterministic polynomial-complete problem. One of the key challenges of scheduling the WAs is to create valid execution sequence. The validity of the execution sequence is...

Full description

Saved in:
Bibliographic Details
Published in:Cluster computing 2020-12, Vol.23 (4), p.3255-3271
Main Authors: Biswas, Tarun, Kuila, Pratyay, Ray, Anjan Kumar
Format: Article
Language:English
Subjects:
Algorithms
Computer Communication Networks
Computer Science
Datasets
Earthquakes
Embedded systems
Energy
Gravitational waves
Incorporation
Iterative methods
Load
Multiple criterion
Objectives
Operating Systems
Optimization
Particle swarm optimization
Polynomials
Processor Architectures
Scheduling
Variance analysis
Workflow software
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Importance of workflow applications (WAs) is expediting in various fields of science and engineering. Scheduling of WAs is a non-deterministic polynomial-complete problem. One of the key challenges of scheduling the WAs is to create valid execution sequence. The validity of the execution sequence is ensured by preserving dependency constraints. Therefore, workflow scheduling algorithms (WSAs) are burning insight to researchers. In this paper, we have proposed a particle swarm optimization based workflow scheduling algorithm to address the problem. Our derived fitness function simultaneously considers several conflicting parameters, makespan, load-balancing, resource-utilization, and speed up ratio. The particle is represented in such a way that it produces a complete solution by preserving the dependency constraints. Moreover, the updated positions of the particles are also ensured to be valid in each iteration. The performance of the proposed work is extensively tested using several scientific WAs. Our simulation results show significant improvements in terms of the considered objectives. The effectiveness of the results is also validated using a statistical hypothesis, Analysis of Variance.
ISSN:1386-7857
1573-7543
DOI:10.1007/s10586-020-03085-3