On infinitely divisible distributions related to the Riemann hypothesis

Takashi Nakamura,Masatoshi Suzuki
2023-06-14
Abstract:For the Riemann zeta-function, we introduce a function such that it is a characteristic function of an infinitely divisible distribution on the real line if and only if the Riemann Hypothesis is true.
Number Theory,Probability
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to provide an equivalent condition for the Riemann Hypothesis (RH) through the relationship between the infinitely divisible distribution and RH. Specifically, the author introduced a function \( g_\zeta(t) \) and proved that the following two propositions are equivalent: 1. **The Riemann Hypothesis holds**. 2. **\( \exp(g_\zeta(t)) \) is the characteristic function of an infinitely divisible distribution on the real line**. Furthermore, if it is assumed that the Riemann Hypothesis holds, the author further described the Lévy measure of the corresponding infinitely divisible distribution and proved that this distribution is a compound Poisson distribution. ### Key Formulas and Concepts - **Riemann ζ - function**: \[ \zeta(s) := \sum_{n = 1}^{\infty}\frac{1}{n^s}, \quad \text{when} \ \Re(s)=\sigma > 1 \] and it can be extended to the entire complex plane \( \mathbb{C} \) by analytic continuation, except for a simple pole at \( s = 1 \). - **Riemann ξ - function**: \[ \xi(s) := \frac{1}{2}s(s - 1)\pi^{-s/2}\Gamma\left(\frac{s}{2}\right)\zeta(s) \] where \( \Gamma(s) \) is the gamma function. The Riemann Hypothesis is equivalent to all non - trivial zeros lying on the critical line \( \Re(s)=\frac{1}{2} \). - **Infinitely divisible distribution**: A probability measure \( \mu \) on \( \mathbb{R} \) is infinitely divisible if for each positive integer \( n \), there exists a probability measure \( \mu_n \) such that \( \mu=\mu_n*\cdots*\mu_n \) (n - fold convolution). - **Lévy - Khintchine formula**: The characteristic function \( \hat{\mu}(t) \) of an infinitely divisible distribution can be expressed as: \[ \hat{\mu}(t)=\exp\left[-\frac{1}{2}at^2 + ibt+\int_{-\infty}^{\infty}\left(e^{it\lambda}-1-\frac{it\lambda}{1 + \lambda^2}\right)\nu(d\lambda)\right] \] where \( (a, b, \nu) \) is a triple and \( \nu \) is the Lévy measure. - **Key function \( g_\zeta(t) \)**: \[ g_\zeta(t):=-4(e^{t/2}+e^{-t/2}-2)+\sum_{n\leq e^t}\Lambda(n)\sqrt{n}(t-\log n)-\frac{t}{2}(\psi(1/4)-\log\pi)+\frac{1}{4}\left(e^{-t/2}\Phi(e^{-2t},2,1/4)-\Phi(1,2,1/4)\right) \] where \( \Lambda(n) \) is the von Mangoldt function, \( \psi(z) \) is the digamma function, and \( \Phi(z, s, a) \) is the Hurwitz - Lerch.