General modelling of layered packet video for multicast systems

摘要

ion Multicasting provides an economic way of distributing information from a sender to a group of receivers located at different sites of the networks. However, heterogeneous networks and variable traffic loads make the multicast more problematic. An efficient solution to the problem is to transmit layered signals, a base layer and several progressive enhancement layers. The local networks simply select the number of layers, according to the capacities. In the paper, a general model of layered packet video is proposed. Two layered compression methods are analysed. The examples show that the general model is suitable to describe layered video, and the model is helpful to design required layers. Layered packet loss rate is suggested as a measure of communication quality requirements. Numerical results show that layered video is more sensitive to the packet dropping than unlayered video and unequal loss protection is more efficient than equal protection.