Abstract: This paper presents Fractional Brownian Motion(FBM) model and applies it in feature analysis of multimedia communication traffic. The multimedia communication traffic under wire network and wireless network is gained through NS simulation experiment. It studies the autocorrelation coefficient, probability characteristic and abruptness of the two kinds of communication traffic. Results prove that the classical FBM model can not characterize all the nature of the multimedia communication traffic.

Key words: Fractional Brownian Motion(FBM) model, streaming media, self-similarity, Hurst parameter

摘要： 提出分形布朗运动(FBM)模型，将其应用于多媒体通信流特性分析。通过NS仿真实验得到有线网络和动态无线网络下的多媒体通信流，研究2种通信流的自相关系数、概率特性和突发性，结果证明典型的FBM模型无法全面刻画多媒体通信流。

关键词: 分形布朗运动模型, 流媒体, 自相似, Hurst参数

CLC Number: 

• TP37