arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2304.00853math.COmath.NT
收藏

关于$\mathbb R$上迭代和与积的一个优于$3/2$的指数

A better than $3/2$ exponent for iterated sums and products over $\mathbb R$

Oliver Roche-Newton

AI总结:

本文证明了实数集上迭代和与积的一个优于$3/2$的指数下界,并据此首次推导出指数严格大于$3/2$的和积估计,还给出了在乘法能量受限时对差集对数的进一步应用。

摘要:

在本文中,我们证明了对所有$A \subset \mathbb R$,以及所有满足额外技术条件的凸函数$f$,不等式\\[ \max \{ |8A-7A|,|5f(A)-4f(A)| \} \gg |A|^{\frac{3}{2} + \frac{1}{54}-o(1)} \\]成立。对数函数满足该技术条件,利用这一事实可推导出和积估计\\[ \max \{ |16A| , |A^{(16)}| \} \gg |A|^{\frac{3}{2} + c}, \\]其中$c>0$。此前,对于任意数量的变量,均不存在指数严格大于$3/2$的$\mathbb R$上的和积估计。此外,$f$的技术条件在多数有趣情形下似乎均被满足,我们给出了进一步的应用。特别地,我们证明了\\[ |AA| \leq K|A| \implies \\,\forall d \in \mathbb R \setminus \{0 \}, \\,\\, |\{(a,b) \in A \times A : a-b=d \}| \ll K^C |A|^{\frac{2}{3}-c'}, \\]其中$c,C>0$为绝对常数。

英文摘要:

In this paper, we prove that the bound \[ \max \{ |8A-7A|,|5f(A)-4f(A)| \} \gg |A|^{\frac{3}{2} + \frac{1}{54}-o(1)} \] holds for all $A \subset \mathbb R$, and for all convex functions $f$ which satisfy an additional technical condition. This technical condition is satisfied by the logarithmic function, and this fact can be used to deduce a sum-product estimate \[ \max \{ |16A| , |A^{(16)}| \} \gg |A|^{\frac{3}{2} + c}, \] for some $c>0$. Previously, no sum-product estimate over $\mathbb R$ with exponent strictly greater than $3/2$ was known for any number of variables. Moreover, the technical condition on $f$ seems to be satisfied for most interesting cases, and we give some further applications. In particular, we show that \[ |AA| \leq K|A| \implies \,\forall d \in \mathbb R \setminus \{0 \}, \,\, |\{(a,b) \in A \times A : a-b=d \}| \ll K^C |A|^{\frac{2}{3}-c'}, \] where $c,C>0$ are absolute constants.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑