The distribution of the overlapping function

Abstract

We consider the set of finite sequences of length n over a finite or countable alphabet C. We consider the function which associate each given sequence with the size of the maximum overlap with a (shifted) copy of itself. We compute the exact distribution and the limiting distribution of this function when the sequence is chosen according to a product measure with marginals identically distributed. We give a point-wise upper bound for the velocity of this convergence. Our results holds for a finite or countable alphabet. The non-parametric distribution is related to the prime decomposition of positive integers. We illustrate with some examples.