arXiv++ Combinatorics

Hickman and Wright proved an $L^2$ restriction estimate for the parabola $Σ$ in $\mathbb{Z}/N\mathbb{Z}$ of the form $$\left(\frac{1}{|Σ|}\sum\limits_{m\inΣ}|\widehat{f}(m)|^2 \right)^{\frac{1}{2}}\leq C_εN^ε\cdot N^{-1}\left(\sum\limits_{x\in (\mathbb{Z}/N\mathbb{Z})^2}|f(x)|^\frac{6}{5}\right)^\frac{5}{6}$$ for all functions $f:(\mathbb{Z}/N\mathbb{Z})^2\rightarrow \mathbb{C}$ and any $ε>0$, and that this bound is sharp when $N$ has a large square factor, and especially for $N = p^2$ for $p$ a prime. In contrast, Mockenhaupt and Tao proved in the special case $N = p$ the stronger estimate $$\left(\frac{1}{|Σ|}\sum\limits_{m\inΣ}|\widehat{f}(m)|^2 \right)^{\frac{1}{2}}\leq C N^{-1}\left(\sum\limits_{x\in (\mathbb{Z}/N\mathbb{Z})^2}|f(x)|^\frac{4}{3}\right)^\frac{3}{4}.$$ We extend the Mockenhaupt-Tao bound to the case of squarefree $N$, proving $$\left(\frac{1}{|Σ|}\sum\limits_{m\inΣ}|\widehat{f}(m)|^2 \right)^{\frac{1}{2}}\leq C_εN^ε\cdot N^{-1}\left(\sum\limits_{x\in (\mathbb{Z}/N\mathbb{Z})^2}|f(x)|^\frac{4}{3}\right)^\frac{3}{4},$$ and discuss applications of this result to uncertainty principles and signal recovery.

We formulate a form of square-root cancellation for the operator which sums a mean-zero function over a hyperplane in $R^d$ for $R$ a possibly noncommutative finite ring. Using an argument of Hart, Iosevich, Koh, and Rudnev, we show that this square-root cancellation occurs when $R$ is a finite field. We then show that this square-root cancellation does not occur over finite rings which are not finite fields. This extends an earlier result of the author to an operator which is not translation-invariant.

Let $R$ be a finite ring (with unit, not necessarily commutative) and define the paraboloid $P = \{(x_1, \dots, x_d)\in R^d|x_d = x_1^2 + \dots + x_{d-1}^2\}.$ Suppose that for a sequence of finite rings of size tending to infinity, the Fourier transform of $P$ satisfies a square-root law of the form $|\hat{P}(χ)|\leq C|R|^{-d}|P|^\frac{1}{2}$ for some fixed constant $C$ (for instance, if $R$ is a finite field, this bound will be satisfied with $C = 1$). Then all but finitely many of the rings are fields. Most of our argument works in greater generality: let $f$ be a polynomial with integer coefficients in $d-1$ variables, with a fixed order of variable multiplications (so that it defines a function $R^{d-1}\rightarrow R$ even when $R$ is noncommutative), and set $V_f = \{(x_1, \dots, x_d)\in R^d|x_d = f(x_1, \dots, x_{d-1})\}$. If (for a sequence of finite rings of size tending to infinity) we have a square root law for the Fourier transform of $V_f$, then all but finitely many of the rings are fields or matrix rings of small dimension. We also describe how our techniques let us see that certain varieties do not satisfy a square root law even over finite fields.