Desarrollo de un modelo predictivo para la estimación del comportamiento de variables en una infraestructura de red

Abstract

The article presents the results found by modeling the variable network packet loss (for a data stream of television signals) and throughput (for a digital video traffic sequence) in a wireless network topology. This is done to estimate its future behavior using autoregressive models integrated with moving averages (ARIMA). All this allows identifying the strength of this type of time series to be applied in internetworking with the purpose of improving its performance. The proposal development is based on the use of the Box- Jenkins methodology. This type of univariate time series is a good candidate to generate behavior predictions in telecommunications networks.