Title :
Pseudorandom bits for constant depth circuits with few arbitrary symmetric gates
Author_Institution :
Div. of Eng. & Appl. Sci., Harvard Univ., Cambridge, MA, USA
Abstract :
We exhibit an explicitly computable ´pseudorandom´ generator stretching l bits into m(l) = lΩ(log l) bits that look random to constant-depth circuits of size m(l) with log m(l) arbitrary symmetric gates (e.g. PARITY, MAJORITY). This improves on a generator by Luby, Velickovic and Wigderson (ISTCS ´93) that achieves the same stretch but only fools circuits of depth 2 with one arbitrary symmetric gate at the top. Our generator fools a strictly richer class of circuits than Nisan´s generator for constant depth circuits (Combinatorica ´91) (but Nisan´s generator has a much bigger stretch). In particular, we conclude that every function computable by uniform poly(n)-size probabilistic constant depth circuits with O(log n) arbitrary symmetric gates is in TIME (2no(1)) This seems to be the richest probabilistic circuit class known to admit a subexponential derandomization. Our generator is obtained by constructing an explicit function f : {0, 1}n → {0, 1} that is very hard on aver-age for constant-depth circuits of size nε·log n with εlog2 n it arbitrary symmetric gates, and plugging it into the Nisan-Wigderson pseudorandom generator construction (FOCS ´88). The proof of the average-case hardness of this function is a modification of arguments by Razborov and Wigderson (IPL ´93), and Hansen and Miltersen (MFCS ´04), and combines Hdstad´s switching lemma (STOC ´86) with a multiparty communication complexity lower bound by Babai, Nisan and Szegedy (STOC ´89).
Keywords :
circuit complexity; communication complexity; logic gates; probability; random number generation; switching circuits; MAJORITY; PARITY; arbitrary symmetric gates; computable function; explicit function; hard problem; multiparty communication complexity; probabilistic circuit; pseudorandom bits; subexponential derandomization; switching lemma; uniform polysize probabilistic constant depth circuits; Circuits; Communication switching; Complexity theory; Computational complexity; Plugs; Polynomials;
Conference_Titel :
Computational Complexity, 2005. Proceedings. Twentieth Annual IEEE Conference on
Print_ISBN :
0-7695-2364-1
DOI :
10.1109/CCC.2005.25