A variational approach to the explicit formula (Q4425993)

The author maintains that theoretical aspects involved in exact explicit formulae seem to be of paramount importance in understanding the fundamental questions on primes, such as the Riemann Hypothesis (RH). He begins his investigation with the Guinand-Weil explicit formula and Weil's positivity criterion which lead to an equivalence between RH and the statement that \({\mathcal T}[f * \overline{f}^{*}]\geq 0\) with equality only if \(f\equiv 0\). Here \(f\) is a complex-valued smooth function which is compactly supported in \({\mathbb R}_{+}\), \(f^{*}(x)={1\over x}f({1\over x})\), \({\mathcal T}[f]\) is the linear functional \[ {\mathcal T}[f] = \int_{0}^{\infty}f(x)\, dx + \int_{0}^{\infty}f^{*}(x)\, dx -\sum_{\nu}W_{\nu}(f) \] with \(\nu\) running over all places \(\{ {\mathbb R},2,3,5,\ldots \}\), and \[ W_{p}(f) = \log p \sum_{m=1}^{\infty}f(p^{m})+f^{*}(p^{m}), \] \[ W_{{\mathbb R}}(f) = (\log 4\pi + \gamma)f(1) + \int_{1}^{\infty}\left(f(x)+f^{*}(x)-{2f(1)\over x}\right){dx\over x-x^{-1}}. \] The author considers the problem of minimization of \({\mathcal T}[f * \overline{f}^{*}]\) in two Hilbert spaces and argues that the assumption of the falsity of RH entails unusual consequences for the behaviour of the extremals. Most of the considerations of this paper also apply to more general \(L\)-series.