Universal Gaussian fluctuations on the discrete Poisson chaos (Q2448709)

According to the authors' classification, a universality result is a mathematical statement implying that the asymptotic behavior of a large random system does not depend on the distribution of its components. Distinguished examples are the central limit theorem, the Donsker theorem and circular laws in random matrix theory. The present paper contains a new class of universality statements involving homogeneous sums based on a sequence of centered independent Poisson random variables. It is proved that homogeneous sums inside a fixed discrete Poisson chaos are universal with respect to normal approximations. One of the main findings is named as ``fourth moment theorem for Poisson sums'' since it provides the conditions of the weak convergence of Poisson sums to normal law together with convergence of their fourth moments. A detailed comparison with previous results and an extended bibliography accompanies the main results.