New Models for Pseudo Self-similar Traac

After measurements on a lan at Bellcore, it is known that data traac is extremely variable on time scales ranging from milliseconds to days. The traac behaves quite diierently to what has been assumed until now; traac sources were generally characterized by short term dependences but characteristics of the measured traac have shown that it is long term dependent. Therefore, new models (such as Fractional Brownian motion, arima processes and Chaotic maps) have been applied. Although they are not easily tractable, one big advantage of these models is that they give a good description of the traac using few parameters. In this paper, we describe a Markov chain emulating self-similarity which is quite easy to manipulate and depends only on two parameters (plus the number of states in the Markov chain). An advantage of using it is that it is possible to re-use the well-known analytical queuing theory techniques developed in the past in order to evaluate network performance. The tests performed on the model are the following: Hurst parameter (by the variances method) and the so-called \visual" test. A method of tting the model to measured data is also given. In addition, considerations about pseudo long range dependences are exposed.