Publication Details

Evolutionary Design of Arbitrarily Large Sorting Networks Using Development

SEKANINA, L.; BIDLO, M. Evolutionary Design of Arbitrarily Large Sorting Networks Using Development. Genetic Programming and Evolvable Machines, 2005, vol. 6, no. 3, p. 319-347. ISSN: 1389-2576.
Czech title
Evoluční návrh libovolně velkých řadicích sítí s využitím developmentu
Type
journal article
Language
English
Authors
URL
Keywords

evolutionary algorithm, development, sorting network, scalability

Abstract

An evolutionary algorithm is combined with an application-specific developmental scheme in order to evolve efficient arbitrarily large sorting networks. First, a small sorting network (that we call the embryo) has to be prepared to solve the trivial instance of a problem. Then the evolved program (the constructor) is applied on the embryo to create a larger sorting network (solving a larger instance of the problem). Then the same constructor is used to create a new instance of the sorting network from the created larger sorting network and so on. The proposed approach allowed us to rediscover the conventional principle of insertion which is traditionally used for constructing large sorting networks. Furthermore, the principle was improved by means of the evolutionary technique. The evolved sorting networks exhibit a lower implementation cost and delay.

Published
2005
Pages
319–347
Journal
Genetic Programming and Evolvable Machines, vol. 6, no. 3, ISSN 1389-2576
BibTeX
@article{BUT46694,
  author="Lukáš {Sekanina} and Michal {Bidlo}",
  title="Evolutionary Design of Arbitrarily Large Sorting Networks Using  Development",
  journal="Genetic Programming and Evolvable Machines",
  year="2005",
  volume="6",
  number="3",
  pages="319--347",
  issn="1389-2576",
  url="https://www.fit.vut.cz/research/publication/7742/"
}
Back to top