Predicting Properties of Congestion Events for a Queueing System With fBm Traffic

In packet networks, congestion events tend to persist, producing large delays and long bursts of consecutive packet loss resulting in perceived performance degradations. The length and rate of these events have a significant effect on network quality of service (QoS). The packet delay resulting from these congestion events also influences QoS. In this paper a technique for predicting these properties of congestion events in the presence of fractional Brownian motion (fBm) traffic is developed.