摘要
为提高在采样环境中Bernaille等的早识别技术对BT(BitTorrent)流量识别的准确性,通过分析BT协议和BT实际传输数据,找出其中BT传输过程前期有效载荷低于70个字节的小包出现的规律,并以此为基础来识别BT流量。实验表明,对比早识别技术,该方法能够提高在采样环境中识别的准确性,克服了早识别技术对BT流中前几个数据包的依赖的不足。
In order to enhance the correctness of Bernaille's Early Identification technology in BT traffic identification under sampling environment,BT protocols and BT actual transmission data are analyzed,rules of the occurrence of small packets whose effective payloads are less than 70 bytes during the early stages of BT transmission procedure are found out,which are taken as the basis for identifying BT traffic.Experiments show that,compared with Early Identification technology,the proposed method can enhance the identification correctness under sampling environment by conquering its lack of dependence on the first few data packets in BT flows.
出处
《计算机应用与软件》
CSCD
北大核心
2012年第1期237-240,共4页
Computer Applications and Software
关键词
P2P
BT
流量识别
小包分析
P2P BT Traffic identification Small packet analysis