Fooling parity tests with parity gates

25Citations
Citations of this article
3Readers
Mendeley users who have this article in their library.
Get full text

Abstract

We study the complexity of computing k-wise independent and ε-biased generators G : {0, l}n → {0,1}m. Specifically, we refer to the complexity of computing G explicitly, i.e. given x ∈ {0,1}n and i ∈ {0, l}log m computing the i-th output bit of G(x). [MNT90] show that constant depth circuits of size poly(n) cannot explicitly compute k-wise independent and ε-biased generators with seed length n ≤ 2log o(1). In this work we show that DLOGTIME-uniform constant depth circuits of size poly(n) with parity gates can explicitly compute k-wise independent and ε-biased generators with seed length n roughly log m ≪ 2log o(1)m. In some cases the seed length of our generators is optimal up to constant factors, and in general up to polynomial factors. To obtain our results, we show a new construction of combinatorial designs, and we also show how to compute, in DLOGTIME-uniform ACo, random walks of length logc n over certain expander graphs of size 2n. © Springer-Verlag 2004.

Cite

CITATION STYLE

APA

Gutfreund, D., & Viola, E. (2004). Fooling parity tests with parity gates. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3122, 381–392. https://doi.org/10.1007/978-3-540-27821-4_34

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