One dimensional stable probability density functions for rational index $\bf 0<\alpha \leq 2$

TL;DR

This paper derives explicit stable probability density functions for rational stability indices using Fox's H-functions, providing new classifications and closed-form expressions for these distributions.

Contribution

It introduces a novel classification scheme for stable laws based on their probability density functions and derives explicit formulas for stable densities with rational indices.

Findings

Explicit formulas for stable densities with rational $eta$

New classification scheme for stable laws

Reproduction of known cases and presentation of new densities

Abstract

Fox's H-function provide a unified and elegant framework to tackle several physical phenomena. We solve the space fractional diffusion equation on the real line equipped with a delta distribution initial condition and identify the corresponding H-function by studying the small $x$ expansion of the solution. The asymptotic expansions near zero and infinity are expressed, for rational values of the index $\alpha$, in terms of a finite series of generalized hypergeometric functions. In $x$-space, the $\alpha=1$ stable law is also derived by solving the anomalous diffusion equation with an appropriately chosen infinitesimal generator for time translations. We propose a new classification scheme of stable laws according to which a stable law is now characterized by a generating probability density function. Knowing this elementary probability density function and bearing in mind the infinitely divisible property we can reconstruct the corresponding stable law. Finally, using the asymptotic behavior of H-function in terms of hypergeometric functions we can compute closed expressions for the probability density functions depending on their parameters $\alpha, \beta, c, \tau $. Known cases are then reproduced and new probability density functions are presented.