Intelligent methods for frame-based analysis of MPEG video models

In this study is presented using a neural networks and learning vector quantization (LVQ) for generating of histogram and a simple Markov chain models of GOP size sequence of MPEG video traffic. The implementation of both models is shown for frame-based analysis of the cell loss probability.