Abstract:We make explicit a theorem of Fromm and Goldmakher [<a class="link-https" data-arxiv-id="1706.03002" href="https://arxiv.org/abs/1706.03002">arXiv:1706.03002</a>], which states that one can improve Burgess' bound for short character sums simply by improving the leading constant in the Pólya-Vinogradov inequality. Towards achieving this, we establish explicit versions of several estimates related to the mean values of real multiplicative functions and the Dickman function.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to explicitly improve the Burgess bound by improving the leading constant in the Pólya - Vinogradov inequality. Specifically:
- **Problem Background**: Given a Dirichlet character \(\chi (\mod q)\), we often need to consider the size of the corresponding character sum \(S_\chi(t)=\sum_{n \leq t} \chi(n)\). For short character sums (i.e., \(t \leq q\)), we have two standard non - trivial estimates: the Pólya - Vinogradov inequality \(S_\chi(t)\ll\sqrt{q}\log q\) and the Burgess bound \(S_\chi(t)\ll t^{1 - \frac{1}{r}}q^{\frac{r + 1}{4r^{2}}+\epsilon}\).
- **Limitations of Existing Methods**: Although the Burgess bound is superior to the Pólya - Vinogradov inequality in some cases, its proof depends on Weil's results and is relatively complex. The standard proof of the Pólya - Vinogradov inequality is relatively simple.
- **Research Objectives**: This paper aims to obtain an explicit improvement of the Burgess bound by improving the leading constant in the Pólya - Vinogradov inequality. Specifically, the author hopes to prove that if the Pólya - Vinogradov inequality can be improved to \(S_\chi(t)\leq(c_1 + o(1))\sqrt{q}\log q\), then for all odd primitive quadratic characters \(\xi (\mod p)\), we have \(S_\xi(t)<ct\), where \(t > p^{\epsilon(c_1,c)}\), and \(\epsilon(c_1,c)=\frac{4\pi c_1}{\delta(c)^{3/2}}+o_t(1)\), \(\delta(c)\leq\frac{2}{7}\).
- **Significance**: This is the first time to show that the Burgess - type results depend on the leading constant in the Pólya - Vinogradov inequality in a meaningful way. Although the current improvement is limited, this result provides a new direction for future research and increases the interest in the relationship between Pólya - Vinogradov and the Burgess bound.
In summary, the core problem of this paper is to explicitly improve the Burgess bound by improving the leading constant in the Pólya - Vinogradov inequality, in order to obtain better estimation results under specific conditions.