English

Efficient quantum protocols for XOR functions

Computational Complexity 2013-07-26 v1

Abstract

We show that for any Boolean function f on {0,1}^n, the bounded-error quantum communication complexity of XOR functions ff\circ \oplus satisfies that Qϵ(f)=O(2d(logf^1,ϵ+lognϵ)log(1/ϵ))Q_\epsilon(f\circ \oplus) = O(2^d (\log\|\hat f\|_{1,\epsilon} + \log \frac{n}{\epsilon}) \log(1/\epsilon)), where d is the F2-degree of f, and f^1,ϵ=ming:fgϵf^1\|\hat f\|_{1,\epsilon} = \min_{g:\|f-g\|_\infty \leq \epsilon} \|\hat f\|_1. This implies that the previous lower bound Qϵ(f)=Ω(logf^1,ϵ)Q_\epsilon(f\circ \oplus) = \Omega(\log\|\hat f\|_{1,\epsilon}) by Lee and Shraibman \cite{LS09} is tight for f with low F2-degree. The result also confirms the quantum version of the Log-rank Conjecture for low-degree XOR functions. In addition, we show that the exact quantum communication complexity satisfies QE(f)=O(2dlogf^0)Q_E(f) = O(2^d \log \|\hat f\|_0), where f^0\|\hat f\|_0 is the number of nonzero Fourier coefficients of f. This matches the previous lower bound QE(f(x,y))=Ω(logrank(Mf))Q_E(f(x,y)) = \Omega(\log rank(M_f)) by Buhrman and de Wolf \cite{BdW01} for low-degree XOR functions.

Keywords

Cite

@article{arxiv.1307.6738,
  title  = {Efficient quantum protocols for XOR functions},
  author = {Shengyu Zhang},
  journal= {arXiv preprint arXiv:1307.6738},
  year   = {2013}
}

Comments

11 pages, no figure

R2 v1 2026-06-22T00:57:47.445Z