Publication Details
Sorting Network Development Using Cellular Automata
Slaný Karel, Ing. (DCSY FIT BUT)
Vašíček Zdeněk, doc. Ing., Ph.D. (DCSY FIT BUT)
Evolutionary algorithm, development, cellular automaton, sorting network, generative encoding.
The sorting network design represents a task that has often been considered as a benchmark for various applications of evolutionary
design and optimization techniques. Although the specific structure of this class of circuits allows to use a simple encoding in combination with additional mechanisms for optimizing the area- and delay-efficiency of designed sorting networks, the design of large sorting networks represents a difficult task. This paper proposes a novel cellular automaton-based approach for the development of specific instances of sorting networks. In order to explore the area of generative cellular automata applied on this specific circuit structures, two di erent encodings are introduced: (1) an absolute encoding and (2) a relative encoding. The abilities of the both techniques are investigated and a comparative study is provided considering a variety of experimental settings.
@INPROCEEDINGS{FITPUB9291,
author = "Michal Bidlo and Karel Slan\'{y} and Zden\v{e}k Va\v{s}\'{i}\v{c}ek",
title = "Sorting Network Development Using Cellular Automata",
pages = "85--96",
booktitle = "Evolvable Systems: From Biology to Hardware",
series = "9th International Conference, ICES 2010, York, UK, September 6-8, 2010, Proceedings, LNCS 6274",
year = 2010,
location = "London, GB",
publisher = "Springer London",
ISBN = "978-3-642-15322-8",
doi = "10.1007/978-3-642-15323-5\_8",
language = "english",
url = "https://www.fit.vut.cz/research/publication/9291"
}