Moment of double exponential sums
Nilanjan Bag, Dwaipayan Mazumder
TL;DR
This paper investigates the $2k$-th moment of double exponential sums in finite fields for the case $s=-1$ and $r=1$, focusing on $\mathcal{S}(\mathcal{M},\mathcal{N},k)=\sum_{m\in\mathcal{M}}\big|\sum_{n\in\mathcal{N}}{\bf e}_p(a\overline{m}n)\big|^{2k}$. The authors combine additive combinatorics (Fourier bias and additive energy) with an ab-shifting technique and Hölder inequalities to derive a nontrivial upper bound for general $\mathcal{M}$ (an interval) and $\mathcal{N}$ (an arbitrary subset):
Abstract
This paper is devoted to finding moments of double exponential sums with monomials over arbitrary sets and intervals in finite fields. The study of such sums dates back to the work of Heath-Brown, who studied such sums in a work on least square-free numbers in an arithmetic progression.
