Convergence of random step lines to Ornstein-Uhlenbeck-type processes (Q1286603)

The authors examine a normalized partial sum process, derived from independent, identically distributed zero mean random variables by progressive Bernoulli thinning of the summands, instead of adding them one by one. The effect of this is to change the timescale on which variation is observed: the limiting process is expressed as an exponentially attenuated Ornstein-Uhlenbeck process. Other models in which the thinned random variables are replaced by further copies of the original random variables are also considered.