Research Article
Scheduling Independent Tasks in Distributed Computing Systems Using NSGA-II
@INPROCEEDINGS{10.4108/eai.7-12-2021.2314587, author={Sarath ambekai}, title={Scheduling Independent Tasks in Distributed Computing Systems Using NSGA-II}, proceedings={Proceedings of the First International Conference on Combinatorial and Optimization, ICCAP 2021, December 7-8 2021, Chennai, India}, publisher={EAI}, proceedings_a={ICCAP}, year={2021}, month={12}, keywords={scheduling multiple objectives optimization problem nsga combinatorial methods}, doi={10.4108/eai.7-12-2021.2314587} }- Sarath ambekai
Year: 2021
Scheduling Independent Tasks in Distributed Computing Systems Using NSGA-II
ICCAP
EAI
DOI: 10.4108/eai.7-12-2021.2314587
Abstract
Task scheduling in heterogeneous multi-processor environments is a complex and challenging issue. This problem in distributed environment is identified to be the Multi-objective Optimization Problems (MOPs), involving the simultaneous fulfillment of several objectives. The complexity of solving MOPs using traditional combinatorial methods are high and hence the optimal solutions can be achieved by using Evolutionary Algorithms. This paper presents a NSGA –II (Non-dominated Sorting Genetic Algorithm) for an efficient scheduling of tasks in a heterogeneous multiprocessorsenvironment. Most of the existing research in this area considered task scheduling with a single objective or bi-objectives only. This scheduling problem is a minimization problem with multiple objectives considering three objectives namely Makespan (Completion Time), Flow Time (Response Time) and Reliability Cost(Fault Tolerance). This method is compared with existing Weighted Sum method (WS). From comparative analysis, NSGA-II provided better performance than WS in twelve different types of heterogeneous environment.