一种基于瑞利分布的VBR视频流的小波模型

本文提出了一种新型的视频业务流模型,以Haar小波的多分辨率分析为基础,在尺度空间和小波空间分别建模,然后通过小波反变换得出仿真业务流.在最"粗"的尺度空间里,我们根据视频流的概率分布特点,采用基于瑞利(Rayleigh)分布的AR模型对尺度系数建模;在各个小波空间里,采用一般的高斯不相关小波模型(WIG,Wavelet Independent Guassian)建模.由于在尺度空间和小波空间针对各自的特点作了不同的处理,本文模型不但能较好拟合复杂业务流在各个时间尺度的概率分布特性,也能拟合其长时相关的特性.另外,在多尺度排队分析(MSQ,MultiScale Queue)的框架下,我们还推导出了基于本文模型的排队分析的理论结果.最后,通过对实际视频业务流数据仿真实验与排队分析验证了本文模型的有效性.

A Rayleigh Distribution Based Wavelet Model for VBR Video Traffic

In this paper,we proposed a new rayleigh distribution based wavelet model for video traffic.With Haar wavelet mutiresolution analysis,we use different models to modeling the scaling coefficients and wavelet coefficients,and by inverse wavelet transform,we can get the whole generated traffic trace.In the coarsest scaling space,we use a rayleigh distribution based AR model to generate the scaling coefficients,for the scaling coefficients are approximately rayleigh distributed.While in the wavelet spaces,we use wavelet independent guassian model to generate the wavelet coefficients.Because the proposed model uses different models in scaling space and wavelet,it can capture the probability distribution and the long-range dependence of the VBR video traffic very well.Moreover,under the multiscale queue (MSQ) framework,we give a theoretic formula of the queuing analysis.Finally,by the video traffic trace-driven simulation and queuing analysis,we show the effectiveness of the proposed model.