Modeling P2P-TV Traffic Using Hidden Markov Models

We propose the use of discrete-time Hidden Markov model (DT-HMM) for representing the P2P-TV traffic. The objective is to develop synthetic traffic generators; or, in other terms, we aim at defining models whose generated synthetic traces are as much as possible “similar” to the real traces. Following the definition presented in [3], a Hidden-Markov model is a doubly embedded stochastic process with an underlying stochastic process that is not observable (it is hidden), but can only be observed through another stochastic process that produces a sequence of observations. In each state of the chain there is a different pattern of bitrate generation. The Hidden-Markov chain is derived by means of a training phase, during which the best fitting with the real trace is looked for. We refer the reader to [3] for a formal presentation of the DT-HMM.