Using Mixed Distribution for Modeling End-to-End Delay Characteristics

An end-to-end round trip time (RTT) is one of the most important communication characteristics for Internet applications. In order to meet the quality of service (QoS) for applications such as streaming services, it is important to understand and predict the delay characteristic accurately. Because characteristics of the packet transmission delay in the Internet vary strongly related to the time-dependent the condition of network. Especially, it is obvious that two or more factors may interfere in observed delay characteristics. We express these delay characteristics by using the mixture of two or more probability distributions (especially normal distribution), not by single distribution. By investigation the factor which infects the delay distribution, we presume the events that occur in network from the delay characteristic obtained by the observation in the end host.