A self-similar traffic generation model based on time

Focus on the obvious self-similarity of much network traffic at present, the paper carries out studies about the generation method of self-similar traffic, and a time-based self-similar traffic generation model is established. By superimposing the Pareto distribution, the model generates the active and idle periods of each source, and the number of the packets generated at any time is obtained though counting the number of the active data source at the moment. The theoretical analysis and results of simulation show that the model has better self-similarity. The model is less complex to compute and easy to apply, and it avoids the obstacles of the traditional ON/OFF model in which the number of packets cannot change with time when self-similar traffic is generated.