Distinguishing photon-count fluctuation characteristics of classical and quantum light sources using non-Markovian processes.

Abstract

Understanding photon-count dynamics is crucial for applications such as quantum random number generators and quantum key distributions, where random and memoryless photon-count arrival processes are essential for performance and security. We investigated the fluctuation dynamics of aggregated photon-count time series for two distinct light sources: a 780-nm laser diode and a spontaneous parametric down-conversion source, using detrended fluctuation analysis (DFA) and the index of dispersion for intervals (IDI). Despite the degradation of photon statistics to Poissonian due to optical losses and detector inefficiencies, our analysis reveals significant long-range dependence at large timescales. The DFA scaling exponents α[over ¯]>0.75 and IDI values c_{k}^{2}>1 at high mean photon count and large aggregation timescales indicate photon-count fluctuations deviated from Poisson model. Our findings underscore the necessity of models beyond Poisson statistics to accurately describe the photon-count dynamics over extended timescales.