Abstract:We establish upper bounds for moments of zeta sums using results on shifted moments of the Riemann zeta function under the Riemann hypothesis.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to provide upper bound estimates for the moments of the ζ - function (Riemann ζ - function). Specifically, the author aims to use the results of the shifted moments of the ζ - function under the Riemann hypothesis to establish upper bounds for the high - order moments of ζ - sums.
### Research Background
In number theory, the study of character sums has been very extensive because they have many important applications in number theory. For example, Harper [3] studied sums of the following forms:
\[
\sum_{n \leq x} n^{it} \quad \text{and} \quad \sum_{n \leq x} \chi(n),
\]
where \( t\in\mathbb{R} \) and \( \chi(n) \) is a non - principal Dirichlet character modulo a large prime \( r \). The first sum is called the zeta sum.
Harper [3] proved that the low - order moments of the zeta sum (and character sums) have the property of "better - than - square - root cancellation". More precisely, he proved that for \( 1\leq x\leq T \) and \( 0\leq k\leq1 \), we have:
\[
\frac{1}{T}\int_{0}^{T}\left|\sum_{n \leq x} n^{it}\right|^{2k}dt\ll(x^{1+(1 - k)\sqrt{\log\log(10L_T)}})^k,
\]
where \( L_T=\min\{x, T/x\} \).
Szabó [9] obtained sharp upper bounds for the shifted moments of Dirichlet L - functions on the critical line under the Generalized Riemann Hypothesis (GRH), and further proved that for a fixed real number \( k > 2 \) and a large integer \( q \), when \( 2\leq Y\leq q^{1/2} \), we have:
\[
\sum_{\chi\in X_q^*}\left|\sum_{n \leq Y} \chi(n)\right|^{2k}\ll_k\phi(q)Y^k(\log Y)^{(k - 1)^2},
\]
where \( X_q^* \) represents the set of primitive Dirichlet characters modulo \( q \), and \( \phi \) is the Euler function.
### Paper's Goal
Inspired by the above work, the author hopes to establish similar upper bounds for the moments of ζ - sums. To this end, the following integral is defined:
\[
S_m(T, Y):=\int_{T}^{2T}\left|\sum_{n \leq Y} n^{-it}\right|^{2m}dt.
\]
The author's main goal is to find an upper bound estimate of \( S_m(T, Y) \). In particular, under the Riemann Hypothesis (RH), the author established the following result:
**Theorem 1.1**: Assume that the Riemann Hypothesis holds. For any real number \( m > 2 \), large real numbers \( T \) and \( Y \) such that \( Y\leq(1-\epsilon)T \) for any \( \epsilon > 0 \), we have:
\[
S_m(T, Y)\ll T Y^m(\log T)^{(m - 1)^2}.
\]
### Conclusion
Through the above research, the author not only improved the results of previous work, but also provided more accurate upper bound estimates for the moments of ζ - sums. This is of great significance for understanding the ζ - function and related problems.