| Abstract: | Video traffic is expected to become increasingly important with the large scale deployment of broadband ISDN. In the literature, it is suggested that smoothing variable bit rate (VBR) video traffic1,2 before transmitting it onto the network would help reduce the probability of cell loss. In this paper we show why deterministic smoothing at the source approximates the minimum achievable loss over the network end-to-end of all possible smoothing schemes. Furthermore, we develop a powerful yet simple analytical technique that can efficiently calculate the loss probability at any point in a network carrying video traffic. We validate our analytical results using traces of actual video segments. The results can be used for admission control and traffic management. We find that in the case of highly correlated traffic such as motion JPEG video, the way to control loss is to ensure that the fraction of time the arrival process exceeds the service process is small. We also find that for such sources increasing the buffer size only marginally decreases the loss. |
