Implementation of Multi-objective Evolutionary Algorithm for Task Scheduling in Heterogeneous Distributed Systems

Yuanlong Chen,Dong Li,Peijun Ma
DOI: https://doi.org/10.4304/jsw.7.6.1367-1374
2012-01-01
Journal of Software
Abstract:This paper presents an effective method for task scheduling in heterogeneous distributed systems. Its objective is to minimize the last task’s finish time and to maximize the system reliability probability. The optimum is carried out through a non-domination sort genetic algorithm. The experimental results based on both randomly generated graphs and the graphs of some real applications showed that, when compared to two well known previous methods, such as heterogeneous earliest finish time (HEFT) algorithm and Critical Path Genetic Algorithm, this algorithm surpasses previous approaches in terms of both last task’s finish time and the system reliability probability.
What problem does this paper attempt to address?