Modelo en Series de Tiempo del Retardo del Cambio de Canal en una Transmisión IPTV

Abstract

ResumenSe presenta el análisis y modelado del comportamiento del retardo generado por una red IPTV a través de series de tiempo, en el momento de realizar una solicitud de cambio de canal. En el estudio se toma en cuenta cada una de las etapas donde el servicio pudiera estar disponible, reduciendo el error promedio del retardo. A través del diseño y simulación de una red IPTV, se desarrolla el análisis del retardo en el cambio de canal para ocho escenarios, desde el usuario hasta la cabecera, considerando el nivel de congestión de la red. La validación de los modelos propuestos se realiza con datos de retardo reales capturados en una red IPTV. Los resultados muestran que cuando la red esta congestionada el modelo que mejor se adapta es el autorregresivo y en caso contrario el de promedios móviles. Palabras AbstractThe analysis and modeling of the delay behavior for channel change request in an IPTV network, through time series are presented. The study takes into account each of the stages where the service would be available, reducing the average delay error. Through the design and simulation of an IPTV network, the delay analysis is developed for eight scenarios, from the user to the header, considering the level of network's congestion. The validation of the proposed models is done with real delay data captured in an IPTV network. The results show that when the network is congested the model that best fits the data is the autoregressive and otherwise is the moving average.

A meta-analysis of channel switching approaches for reducing zapping delay in internet protocol television

“…The selected eight kinds of literature simulated there a method for more than one user hence giving more credibility to the result reported When comparing the network approach to the other three approaches, the forest plot in Figure 3 indicates that the network approach performed better in terms of reducing zapping delay. The recently proposed network approaches include a time series model [19] that takes into account each of the stages where the service could be available, reducing the average delay error from the client to the headend. The results show that when the network is congested the model that best fits the data is the autoregressive Furthermore, to reduce zapping delay, [2] suggested using a hybrid delivery system that combines multicast and unicast-enabled services.…”

“…They used channel reordering and preload policy to classify genre for channels to shorten seek distance and zapping time. Sarni et al [19] further presented an approach at the network level called the channel switching scenario in Multicast IPTV Network. Their method switches between channels by sending a join message to the desired channel before leaving the one being watched.…”