On the mean-square of the Riemann zeta-function on the critical line (Q1117971): Difference between revisions

The central theme of this paper is the analogy between the functions E(T) (the error term in the mean square formula for \(\zeta (1/2+it))\) and \(\Delta\) (x) (the error term in the divisor problem). For instance, the authors obtain for E(T) exact analogues of the best known \(\Omega_+\)- and \(\Omega_-\)-estimates for \(\Delta\) (x). The similarity between Voronoi's formula for \(\Delta\) (x) and Atkinson's formula for E(T) is utilized, but there are some difficulties owing to the more complicated structure of Atkinson's formula. To overcome these, the average of E(t) over the interval [2,T] is introduced and written approximately as an absolutely convergent series. For the related error term \(G(T)=\int^{T}_{2}E(t) dt-\pi T,\) the following results are proved: \[ \int^{T}_{2}G^ 2(t)dt=BT^{5/2}+O(T^ 2) \] (B is a numerical constant) and \(G(T)=\Omega_{\pm}(T^{3/4})\).