摘要： 通过实验分析了国内流行的即时通信软件——腾讯QQ的流量特征以及其语音会话的流量特征，应用净荷深度检测(DPI)和智能会话关联(ISA)技术来识别腾讯语音通话流量，设计了腾讯语音业务流量的识别模型和算法。模型和算法的准确性、可扩展性和健壮性在电信运营商IP骨干网中得到了验证。
关键词:
净荷深度检测,
流量识别,
VoIP识别,
会话关联
Abstract: This paper aims at voice traffic characteristics of Tencent QQ, which is the currently popular Internet instant communication software. It deeply focuses on how to effectively identify QQ voice traffic based on deep packet inspection(DPI) and intelligent session association(ISA) technology, then presents a model of identification and related algorithm. The voice identification system and the accuracy, scalibity, robustness of the model and algorithm are validated in core network of China Telecom.
Key words:
deep packet inspection(DPI),
traffic identification,
VoIP identification,
session association
中图分类号:
王 攀;金 婷;张顺颐;陈雪娇;李 薇. 基于智能会话关联的腾讯语音流量识别算法[J]. 计算机工程, 2007, 33(17): 137-139.
WANG Pan; JIN Ting; ZHANG Shun-yi; CHEN Xue-jiao; LI Wei. Algorithm of Tencent’s Voice Traffic Identification Based on Intelligent Session Association[J]. Computer Engineering, 2007, 33(17): 137-139.