Under the auspices of the Computational Complexity Foundation (CCF)

REPORTS > DETAIL:

Paper:

TR04-088 | 12th October 2004 00:00

Fooling Parity Tests with Parity Gates

TR04-088
Authors: Emanuele Viola, Dan Gutfreund
Publication: 22nd October 2004 16:57
Keywords:

Abstract:

We study the complexity of computing $k$-wise independent and
$\epsilon$-biased generators $G : \{0,1\}^n \to \{0,1\}^m$.
Specifically, we refer to the complexity of computing $G$ \emph{explicitly}, i.e.
given $x \in \{0,1\}^n$ and $i \in \{0,1\}^{\log m}$, computing the $i$-th output bit of $G(x)$.
Mansour, Nisan and Tiwari (1990) show that constant depth circuits of size poly(n) (i.e. $AC^0$) cannot explicitly compute
$k$-wise independent and $\epsilon$-biased generators with seed length $n = 2^{\log^{o(1)} m}$.

In this work we show that DLOGTIME-uniform constant depth circuits
of size poly(n) \emph{with parity gates} can explicitly compute
$k$-wise independent and $\eps$-biased generators with seed length $n = (\poly) \log m$.
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 $AC^0$, random walks
of length $log^c n$ over certain expander graphs of size $2^n$.

ISSN 1433-8092 | Imprint