Bidirectional Statistical Feature Extraction Based on Time Window for Tor Flow Classification

The anonymous system Tor uses an asymmetric algorithm to protect the content of communications, allowing criminals conceal their identities and hide tracks. This malicious usage brings serious security threats public social stability. Statistical analysis traffic flows can effectively identify classify flow. However, few features be extracted from traffic, which have a weak representational ability, making it challenging combat cybercrime in real-time effectively. Extracting utilizing more accurate is key point improving detection performance traffic. In this paper, we design efficient identification scheme for based on time window method bidirectional statistical characteristics. divide network by sliding then calculate relative entropy We adopt sequential pattern mining extract application types Finally, extensive experiments are carried out UNB dataset (ISCXTor2016) validate our proposal’s effectiveness property. experiment results show that proposed detect flow with accuracy 93.5% 91%, respectively, speed processing classifying single 0.05 s, superior state-of-the-art methods.