Superfast Fourier transform using QTT approximation

Sergey Dolgov, Boris N. Khoromskij, and Dmitry Savostyanov

Contact the author: Please use for correspondence this email.
Submission date: 26. Apr. 2011 (revised version: March 2012)
Pages: 44
published in: The journal of Fourier analysis and applications, 18 (2012) 5, p. 915-953 
DOI number (of the published article): 10.1007/s00041-012-9227-4
Bibtex
MSC-Numbers: 15A23, 15A69, 65F99, 65T50
Keywords and phrases: Multidimensional arrays, QTT, FFT, sine/cosine transform, convolution, sparse Fourier transform, high-dimensional problems, tensor train format, Fourier transform, quantum Fourier transform
Download full preprint: PDF (2324 kB)

Abstract:
We propose Fourier transform algorithms using QTT format for data-sparse approximate representation of one– and multi-dimensional vectors (m–tensors). Although the Fourier matrix itself does not have a low-rank QTT representation, it can be efficiently applied to a vector in the QTT format exploiting the multilevel structure of the Cooley-Tukey algorithm. The m–dimensional Fourier transform of an n ×… × n vector with n = 2^d has (md²R³) complexity, where R is the maximum QTT–rank of input, output and all intermediate vectors in the procedure. For the vectors with moderate R and large n and m the proposed algorithm outperforms the (n^m log n) fast Fourier transform (FFT) algorithm and has asymptotically the same log-squared complexity as the superfast quantum Fourier transform (QFT) algorithm. By numerical experiments we demonstrate the examples of problems for which the use of QTT format relaxes the grid size constrains and allows the high-resolution computations of Fourier images and convolutions in higher dimensions without the “curse of dimensionality”. We compare the proposed method with Sparse Fourier transform algorithms and show that our approach is competitive for signals with small number of randomly distributed frequencies and signals with limited bandwidth.