Identification of cryptographically strong and weak pseudorandom bit generators

0Citations
Citations of this article
3Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Cryptographically strong Psuedo Random Bit Generators (PRBGs) are used extensively for confidentiality of information. If these PRBGs are weak then the security of information is at stack. In this paper we are giving a model which will use confidence interval and multidimentional scaling for their identification. This statistical model can be used to estimate the strength of a new PRBGs. © 2008 Springer Berlin Heidelberg.

Cite

CITATION STYLE

APA

Verma, N., Mishra, P. R., & Pandey, G. (2008). Identification of cryptographically strong and weak pseudorandom bit generators. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5352 LNCS, pp. 154–155). https://doi.org/10.1007/978-3-540-89862-7_13

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