English

On additive shifts of multiplicative almost-subgroups in finite fields

Number Theory 2015-07-21 v1

Abstract

We prove that for sets A,B,CFpA, B, C \subset \mathbb{F}_p with A=B=Cp|A|=|B|=|C| \leq \sqrt{p} and a fixed 0dFp0 \neq d \in \mathbb{F}_p holds max(AB,(A+d)C)A1+1/26. \max(|AB|, |(A+d)C|) \gg|A|^{1+1/26}. In particular, A(A+1)A1+1/26 |A(A+1)| \gg |A|^{1 + 1/26} and max(AA,(A+1)(A+1))A1+1/26. \max(|AA|, |(A+1)(A+1)|) \gg |A|^{1 + 1/26}. The first estimate improves the bound by Roche-Newton and Jones. In the general case of a field of order q=pmq = p^m we obtain similar estimates with the exponent 1+1/559+o(1)1+1/559 + o(1) under the condition that ABAB does not have large intersection with any subfield coset, answering a question of Shparlinski. Finally, we prove the estimate xFqψ(xn)q72δ28n2+2δ28 \left| \sum_{x \in \mathbb{F}_q} \psi(x^n) \right| \ll q^{\frac{7 - 2\delta_2}{8}}n^{\frac{2+2\delta_2}{8}} for Gauss sums over Fq\mathbb{F}_q, where ψ\psi is a non-trivial additive character and δ2=1/56+o(1)\delta_2 = 1/56 + o(1). The estimate gives an improvement over the classical Weil bound when q1/2n=o(q29/57+o(1))q^{1/2} \ll n = o\left( q^{29/57 + o(1)} \right).

Keywords

Cite

@article{arxiv.1507.05548,
  title  = {On additive shifts of multiplicative almost-subgroups in finite fields},
  author = {Dmitrii Zhelezov},
  journal= {arXiv preprint arXiv:1507.05548},
  year   = {2015}
}