自相似模型仿真ATM网络VBR视频源研究

摘要

ATM is the basic transmission and switching technology used in the further B-ISDN network. This paper proposes an efficient method for video traffic modeling, called long-range dependence analysis and selfsimilar traffic modeling. Using FGN process, one of self-similar processes and making improvements on the FGN algorithm to model both SLD and LRD of the empirical sequence autocorrelation structure, we present simulation results of synthetic data and compare with results based on empirical video sequence. Basing on statistical analysis and graphical tests of the data got from the empirical video sequence, we come to the conclusion that self-similar process is ideal for modeling VBR video traffic.%仿真过程采用了一种新的异步传输模式(ATM)视频源建模方法，即长期相关性分析和自相似模型方法，使用了自相似模型中的分形高斯噪声(FGN)模型.FGN过程的产生采用了Hosking算法和反变换法.在仿真具体实现过程中改进了现有FGN模型仿真算法的不足，从而使产生的过程既充分考虑到实际可变比特率(VBR)视频源序列的长期相关性，又兼顾了序列的短期相关性，同时给出了仿真结果与实际源统计特性的比较和验证.比较结果表明，自相似模型是一种理想的VBR视频源模型.