Soft computing for adaptative multimedia applications

Although IPv6 and other technologies have improved the ways we deal with QoS (Quality of Service) in internet, it is still a problem to deal with high quality distributed multimedia applications running over heterogeneous networks.

This heterogeneity can be spatial, or temporal. Spatial in the sense that we may know in advance that each specific network segment, can (or cannot) provide a defined QoS. In this case, manual tuning is a costly, but valid option. Through tuning you may use transcoders, flow aggregators/mixers, or just adjust some parameters.

But in general, conditions in the network may vary due to temporal causes, congestion, dynamic routing, high load, and may other reasons. This changes in the underlying infrastructure, cannot be known in advance. And it is not always possible to have a minimum QoS guarantee. In this cases, the best way to get optimum quality in your application is making it adaptative.

My proposal is using a neural network to control parameters for adaptative multimedia applications over heterogeneus networks and/or networks without guaranteed QoS.


Para mantener la alta calidad de los contenidos, debes acceder para dejar un comentario
Las cookies nos ayudan a ofrecer nuestros servicios. Al utilizar nuestros servicios, aceptas el uso de cookies. Más información