P2P Traffic Identification Based on Fixed-Length Packets
As the key point of network traffic identification, the detection of encrypted peer-to-peer (P2P) file-sharing traffic is still a great challenge. In this paper, a P2P traffic identification method based on fixedlength packets is proposed from three aspects which are suitable-traffic collection, pattern extraction and pattern matching. The pattern extraction is abstracted as an extended longest common subsequence problem and the pattern matching is modeled as a simplified episode matching problem. Efforts are taken to increase the efficiency and decrease the false-positive ratio of identification. The preliminary evaluation results show that our method is effective for identifying the encrypted P2P file-sharing traffic which contains the fixedlength packets.
Peer-to-peer file-sharing application traffic identification encrypted traffic extended longest common subsequence simplified episode matching
Ruhui Zhang Ye Du Jun Liu Xing Wang Zhonglan Yuan
School of computer and information technology, Beijing jiaotong university, Beijing
国际会议
2010 Cross-Strait Conference on Information Science and Technology(2010 海峡两岸信息科学与技术学术交流会)
秦皇岛
英文
761-765
2010-07-09（万方平台首次上网日期，不代表论文的发表时间）