The mean square of the product of the Riemann zeta-function with Dirichlet polynomials (Q2398715): Difference between revisions

The main object of study of this interesting paper is the integral \[ I: = \int_{-\infty}^\infty |\zeta(1/2+it)|^2|A(1/2+it)|^2\phi(t/T)\,dt, \] where \(\phi(x)\) is a smooth function supported in \([1,2]\), and \[ A(s) := \sum_{n\leq T^\theta}a_nn^{-s},\quad 0 < \theta < 1,\quad a_n \ll n^\varepsilon. \] In their main result, Theorem 1, the authors prove the asymptotic formula for \(I\) with the error term which is \[ \ll_\varepsilon T^{3/20+\varepsilon}N^{23/20} + T^{1/3+\varepsilon}, \] with \(N = T^\theta, \theta < 1/2 +\delta, \delta = 1/66\). The main tool in the proof of Theorem 1 is a forthcoming estimate for trilinear forms of Kloosterman fractions. This estimate improves a result of \textit{W. Duke} et al. [Invent. Math. 128, No. 1, 23--43 (1997; Zbl 0873.11050)]. It is crucially important to allow \(\theta\) in Theorem 1 to be as large as possible. For example, if we could take \(\theta = 1 - \varepsilon\), then the Lindelöf Hypothesis (\(\zeta(1/2+it) \ll_\varepsilon |t|^\varepsilon\)) would follow. If one assumes a general estimate for trilinear forms of Kloosterman fractions, namely the authors' (1.3), then Theorem 1 can be replaced by their Theorem 2. As an application of Theorem 1 the authors obtain an upper bound of the correct order of magnitude for the third moment of the Riemann zeta-function. This new bound is \[ \int_0^T|\zeta(1/2+it)|^3\,dt \ll T(\log T)^{9/4}, \] and the lower bound of the same form has been known for a long time; see e.g., [\textit{K. Ramachandra}, Lectures on the mean-value and Omega-theorems for the Riemann Zeta-function. Berlin: Springer-Verlag; Bombay: Tata Institute of Fundamental Research (1995; Zbl 0845.11003)]. Finally, their Theorem 3 is devoted to the study (\(\phi(t/T)\) is missing in their (1.5)) of the integral \[ J := \int_{-\infty}^\infty |\zeta(1/2+it)|^2|A(1/2+it)|^2|B(1/2+it)|^2\phi(t/T)\,dt, \] where \[ A(s):= \sum_{n\leq N}\alpha_n n^{-s},\quad B(s) := \sum_{k\leq K}\beta_k k^{-s}, \quad \alpha_n \ll_\varepsilon n^\varepsilon, \beta_k \ll_\varepsilon k^\varepsilon. \]