English

Discrete Fourier restriction via efficient congruencing: basic principles

Classical Analysis and ODEs 2024-07-01 v1 Number Theory

Abstract

We show that whenever s>k(k+1)s>k(k+1), then for any complex sequence (an)nZ(\mathfrak a_n)_{n\in \mathbb Z}, one has [0,1)knNane(α1n++αknk)2sdαNsk(k+1)/2(nNan2)s.\int_{[0,1)^k}\left| \sum_{|n|\le N}\mathfrak a_ne(\alpha_1n+\ldots +\alpha_kn^k) \right|^{2s}\,{\rm d}{\mathbf \alpha}\ll N^{s-k(k+1)/2}\biggl( \sum_{|n|\le N}|\mathfrak a_n|^2\biggr)^s. Bounds for the constant in the associated periodic Strichartz inequality from L2sL^{2s} to l2l^2 of the conjectured order of magnitude follow, and likewise for the constant in the discrete Fourier restriction problem from l2l^2 to LsL^{s'}, where s=2s/(2s1)s'=2s/(2s-1). These bounds are obtained by generalising the efficient congruencing method from Vinogradov's mean value theorem to the present setting, introducing tools of wider application into the subject.

Keywords

Cite

@article{arxiv.1508.05329,
  title  = {Discrete Fourier restriction via efficient congruencing: basic principles},
  author = {Trevor D. Wooley},
  journal= {arXiv preprint arXiv:1508.05329},
  year   = {2024}
}

Comments

37 pages

R2 v1 2026-06-22T10:38:57.778Z