Removing logarithms from Poisson process error bounds. (Q1877394)

A new approximation theorem for estimating the error in approximating the whole distribution of a finite-point process by a suitable Poisson process is presented. A point process is defined as a random element taking values in the space of finite configurations on a compact metric space. The Wasserstein metric is used for estimating the errors in Poisson approximation of point processes. The approach is based on finding a new smoothness estimate and the Palm process approximation. In many cases, the new bounds remain constant as the mean of the process increases, in contrast to previous results which, at best, increase logarithmically with the mean. Applications are given to Bernoulli trials, a randomly shifted Bernoulli process, and to networks of queues. In these applications the bounds are independent of time and space, only dependent on parameters of the system under consideration.