Local moment conjecture for exponential sums

About 5 years old · traced to

For d≥2d\geq 2, let Sa,d(x,N)=∑n=1Nane(x1n+⋯+xdnd)S_{a,d}(x,N)=\sum_{n=1}^N a_n e(x_1n+\cdots+x_dn^d) and let ρd\rho_d be

ρd={3d2−44,d even,frac3d2−34,d odd.\rho_d=\begin{cases}\frac{3d^2-4}{4},&d\text{ even},\\frac{3d^2-3}{4},&d\text{ odd}.\end{cases}

Local moment conjecture. For every j≥0j\geq 0,

2jd+12∫[0,2−j]d∣Sa,d(x,N)∣p dx≲ϵ∥a∥ℓ2p{Nϵ,0<p≤ρd,Np−ρd2+ϵ,p>ρd.2^{j\frac{d+1}{2}}\int_{[0,2^{-j}]^d}|S_{a,d}(x,N)|^p\,dx\lesssim_\epsilon\|a\|_{\ell^2}^p\begin{cases}N^\epsilon,&0<p\leq\rho_d,\\N^{\frac{p-\rho_d}{2}+\epsilon},&p>\rho_d\end{cases}.

This local estimate is intended as a strategy for the Fourier-decay conjecture and is sharp under the constructive-interference examples discussed in the paper; it remains open in general.

References

Primary source

Ciprian Demeter and Bartosz Langowski, “Restriction of exponential sums to hypersurfaces”, arXiv:2104.11367 (2021).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.