摘要
为了提高Skype语音流的检测效率,在分析了目前识别技术的基础上,得出了"基于UDP的Skype通信是系统中的薄弱环节"等看法。由此深入分析了Skype流量特征和负载特征,提出了一种基于UDP的Skype语音流检测算法SUFC,该算法综合利用了UDP分组中的特征和VoIP的统计特征,算法复杂性较低,便于在线实施。给出了能够实时在线检测Skype语音流的原型系统的关键技术。试验结果表明,基于SUFC算法实现的原型系统具有效率高和差错率小的特点。
To improve the efficiency of Skype voice flow identification, current techniques were analyzed to get a viewpoint that the Skype communication based on UDP was a weak part. Therefore, the characteristics of flow Skype and load were analyzed thoroughly; and an algorithm named SUFC (Skype-UDP-Flow- Characteristics-based) was proposed to identify the flow of Skype based on UDP. SUFC utilized both the UDP packet characteristics and the statistic characteristics of VolP session, which had low complexity and could be implemented online. Some key techniques for real-time Skype voice flow online identification in the prototype system of SUFC were discussed. The results of experiments show that the prototype system using SUFC algorithm is of high efficiency and low error rate.
出处
《解放军理工大学学报(自然科学版)》
EI
2008年第5期507-511,共5页
Journal of PLA University of Science and Technology(Natural Science Edition)
基金
国家863计划资助项目(2007AA01Z418)
关键词
网络测量
SKYPE
检测
流
network measurement
skype
identification
flow