Investigators
Z. Obradovic, R. Drossu, and C. Raghavendra, T.V. Lakshman
Problem
Accurate traffic prediction can be used to optimally smooth delay sensitive traffic and increase multiplexing gain in asynchronous transfer mode (ATM) networks.
The objectives of this work are to investigate applicability of neural network techniques for prediction of either the following or several following frame sizes using the information of previous frame sizes.
Results
Neural network models are considered for two important types of video sequences - video teleconferencing and entertainment video. An off-line learning method is suggested for simple traffic (drossu94) and an on-line learning method for complex one (drossu96book).
Simulation studies of cell losses in an ATM multiplexer using recorded variable-bit-rate coded video teleconference data indicate reasonably good predictions for buffer delays between 0.5 and 5 ms.