加密MANET端到端流的推断与识别研究

根据数据流的帧发送时序特性, 提出一种基于时序匹配和关联分析的端到端流推断算法。采用时序匹配、关联分析、流追踪与回溯等机制, 分析与追踪探测区域内的所有端到端流。该算法能够在无须解密报文的前提下进行流量分析, 推断出端到端流的源、目的节点、流路径、流开始时间和流持续时长等信息。仿真实验表明, 算法可以有效识别网络中的端到端流。

End-to-end traffic inference of encrypted MANET

According to the characteristics of the frame sending time sequences, this paper proposed a time sequence matching and correlation analysis based end-to-end traffic analysis algorithm. Adopting time sequence matching, correlation analysis, and flow tracer to infer and trace all the end-to-end flows in the network, the algorithm could infer the traffic pattern without decrypting the frames. Simulation experiment results verify the effectiveness of the proposed method.