A hierarchical classification approach for tor anonymous traffic

Tor is an anonymous communication system that can protect our privacy, but it also provides a haven for criminals to avoid network tracing. Therefore, anonymous traffic analysis and classification is an important part of maintaining network security. Existing Tor traffic classification methods require a large number of labeled data, and the classification accuracy rate is not satisfied for practical using. This paper presents a hierarchical classification approach for Tor anonymous traffic. An improved decision tree algorithm (Tor-IDT) is used to identify the Tor anonymous traffic from the mixed traffic, and then the Tri-Training algorithm is used to segment the identified anonymous traffic at the application level. Experiments show that the recognition rate of Tor anonymous traffic is more than 99%, the accuracy of classification can reach 94%, and less labeled data is needed, which proves that this hierarchical classification algorithm has wider applicability and higher classification accuracy.