Performance evaluation of video signals at burst level in an ATM environment

This paper focuses on the performance evaluation on two methods namely the leaky bucket and fast buffer reservation at the burst level for variable bit rate video signals. There has been a considerable amount of work on traffic shaping methods such as the leaky bucket. Much of this work has concentrated on the cell level instead of the burst level to generate randomly tokens at the cell level. Most authors suggested that error handling is strongly recommended to be done on an end-to-end call basis rather than link-by-link, as the latter method involves extensive protocol validation and recovery which is a significant timing factor in the asynchronous transfer mode (ATM). This scheme is offset by retransmitting control data from the source instead of from immediate nodes along the call set up path. We show that even if the overall cell discard probability, throughput and lost burst using a fast buffer reservation scheme are not much better than the leaky buffer method, the procedures of error recovery due to lost cells are simpler and more efficient