Extreme Value Statistics in Silicon Photonics

Abstract

L-shape probability distributions are extremely non-Gaussian functions that have been surprisingly successful in describing the occurrence of extreme events ranging from stock market crashes, natural disasters, structure of biological systems, fractals, and optical rogue waves. We show that fluctuations in stimulated Raman scattering, as well as in coherent anti-Stokes Raman scattering, in silicon can follow extreme value statistics and provide mathematical insight into the origin of this behavior.