It is well known that any nondeterministic finite automata over a unary alphabet can be represented in a certain normal form called the Chrobak normal form [1]. We present a very simple conversion procedure working in O(n 3) time. Then we extend the algorithm to improve two trade-offs concerning conversions between different representations of unary regular languages. Given an n-state NFA, we are able to find a regular expression of size O(n2/log2n) describing the same language (which improves the previously known O(n2) size bound [8]) and a context-free grammar in Chomsky normal form with O(√n log n) nonterminals (which improves the previously known O(n2/3) bound [3]). © 2011 Springer-Verlag.
CITATION STYLE
Gawrychowski, P. (2011). Chrobak normal form revisited, with applications. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 6807 LNCS, pp. 142–153). https://doi.org/10.1007/978-3-642-22256-6_14
Mendeley helps you to discover research relevant for your work.