On the discrete kolmogorov–wiener filter for the one-point prediction of exponentially smoothed heavy-tail processes

Abstract

The prediction of telecommunication traffic is an important problem for telecommunications and cyber security, see a detailed description in [1]. There are a plenty of different (and rather sophisticated) approaches to traffic prediction, see [1]. The telecommunication traffic is considered to be stationary random process in a couple of models, and, as is known, such a simple algorithm as the Kolmogorov–Wiener filter may be applied to prediction of stationary processes. So, it is of interest to investigate the possibility of the Kolmogorov–Wiener filter application to heavy-tail process prediction, because traffic in telecommunication systems with data packet transfer in considered to be a heavy-tail random process, see [2,3]. Out previous paper [4] is devoted to the corresponding problem.