An adaptation scheme for real-time video streaming

The transmission of digital video on the Internet has become popular recently. However, the transmission bandwidth may be limited in many situations. Therefore, a bandwidth adaptation mechanism, which is able to dynamically select suitable bandwidth in real-time , is preferred for video transmission. In this paper, we make use of the naive Bayes classifier to build such an adaptation scheme. The decision parameters of this classifier include the packet loss rate, the network instant available bandwidth, and the transmission bit rate. The objective of this designed adaptation scheme aims to reduce the transmission packet loss rate and to select the suitable bandwidth for transmission. In this paper, we describe our design and illustrate the effectiveness of the design via experimental simulations in an ADSL environment.