The Study of Streaming Traffic behavior

Abstract

The use of computer network has no longer been limited to sending and/or receiving files, web
browses, or e-mails; it has been designed and developed continuously to serve the users' requirements.
Nowadays, teleconference and entertainment online have been widely used; most of those
applications are streaming traffics which would play an important role for computer network in the
near future. Therefore, the understanding of the streaming traffic’s working process is very important.
It would be useful for better designing of the computer network system. This research studies the
characteristic of streaming traffic such as on demand traffic and real-time live broadcast traffic. The
various captured traffics have been analyzed and we could make a conclusion of the difference
between on demand traffic and live streaming traffic behavior. And we also find the relationship
between the encoding rate and the streaming traffic behavior; as when the source file is encoded with
higher encoding rate, the average of the packet size increases and the packet time interval decreases.
Moreover, this research also develops the streaming traffic model for both on demand traffic and realtime
live broadcast traffic. The developed traffic model could generate traffic similar to the real traffic
as shown in the experimental results.