Succinct descriptions of regular languages with binary ⊕-NFAs

Abstract

Champarnaud [1] analyzed the number of states obtained from a binary ⊕-NFA during the subset construction. We extend this work to an experimental analysis of the size of the minimal DFAs obtained from binary ⊕-NFAs. We then consider the number of distinct languages accepted by binary ⊕-NFAs, and compare that to Domaratzki's results [2] for (traditional) binary NFAs. We also show that there are certain regular languages which are accepted by succinct ⊕-NFAs, but for which no succinct traditional NFA exists. © Springer-Verlag Berlin Heidelberg 2003.