Sparse Fourier Transform in Any Constant Dimension with Nearly-Optimal Sample Complexity in Sublinear Time
Abstract
We consider the problem of computing a -sparse approximation to the Fourier transform of a length signal. Our main result is a randomized algorithm for computing such an approximation (i.e. achieving the sparse recovery guarantees using Fourier measurements) using samples of the signal in time domain that runs in time , where is the dimensionality of the Fourier transform. The sample complexity matches the lower bound of for non-adaptive algorithms due to \cite{DIPW} for any for a constant up to an factor. Prior to our work a result with comparable sample complexity and sublinear runtime was known for the Fourier transform on the line \cite{IKP}, but for any dimension previously known techniques either suffered from a polylogarithmic factor loss in sample complexity or required runtime.
Cite
@article{arxiv.1604.00845,
title = {Sparse Fourier Transform in Any Constant Dimension with Nearly-Optimal Sample Complexity in Sublinear Time},
author = {Michael Kapralov},
journal= {arXiv preprint arXiv:1604.00845},
year = {2016}
}