| Summary: | Most of the traffic on the Internet is encrypted traffic, and the detection of encrypted traffic is the current difficulty, because the internal features of the data are destroyed after encryption, and it is difficult to detect. Most of the existing detection of encrypted traffic is based on the external features of encrypted traffic, which requires the extraction of full-cycle information of traffic, and has poor real-time performance. Therefore, based on the internal features of encrypted traffic, this paper proposes a Key Feature Fusion Detection (KFFD) method based on generative adversarial network, which restores the destroyed internal features by encryption key and generative adversarial network, and then improves the internal feature recognition effect of encrypted traffic. Experiments using the Kaggle dataset show that the KFFD method can improve the detection performance of encrypted traffic to a certain extent.
|
|---|