Counting equivalent linear finite transducers using a canonical form

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

Abstract

The notion of linear finite transducer (LFT) plays a crucial role in a family of cryptosystems introduced in the 80's and 90's. However, as far as we know, no study was ever conducted to count and enumerate these transducers, which is essential to verify if the size of the key space, of the aforementioned systems, is large enough to prevent an exhaustive search attack. In this paper, we determine the cardinal of the equivalence classes on the set of the LFTs with a given size. This result is sufficient to get an approximate value, by random sampling, for the number of non-equivalent injective LFTs, and subsequently for the size of the key space. We introduce a notion of canonical LFT, give a method to verify if two LFTs are equivalent, and prove that every LFT has exactly one equivalent canonical LFT. We then show how this canonical LFT allows us to calculate the size of each equivalence class on the set of the LFTs with the same number of states. © 2014 Springer International Publishing Switzerland.

Cite

CITATION STYLE

APA

Amorim, I., Machiavelo, A., & Reis, R. (2014). Counting equivalent linear finite transducers using a canonical form. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8587 LNCS, pp. 70–83). Springer Verlag. https://doi.org/10.1007/978-3-319-08846-4_5

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