Cellular automaton as sorting network generator using instruction-based development

1Citations
Citations of this article
3Readers
Mendeley users who have this article in their library.
Get full text

Abstract

A new cellular automaton-based approach allowing to generate sorting networks is presented. Since the traditional table-based transition function in this case involves excessive number of rules, a program-based representation of the transition function is applied. The sorting networks are encoded by the cell states and generated during the cellular automaton development. The obtained results are compared with our previous approaches utilizing cellular automata. © 2012 Springer-Verlag Berlin Heidelberg.

Cite

CITATION STYLE

APA

Bidlo, M., & Vasicek, Z. (2012). Cellular automaton as sorting network generator using instruction-based development. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7495 LNCS, pp. 214–223). Springer Verlag. https://doi.org/10.1007/978-3-642-33350-7_22

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free