Exploration on Malicious Encrypted Traffic Classification Based on Deep Learning

In recent years, encrypted traffic has become the main means of network attacks, especially the use of encrypted traffic for network attacks is becoming increasingly common. In real life, some malicious encrypted traffic is attacked through domain name spoofing or domain hijacking, while in the internet, these malicious encrypted traffic is disguised through various means, causing the purpose of network attacks. To address this issue, the article introduced a malicious encrypted traffic classification method based on deep learning (DL). The article first classified encrypted traffic, and then used DL methods to classify it. Finally, the model was trained and tested using multiple datasets. The experimental results showed that the proposed model was superior to other algorithms in accuracy, up to 91.4%, and also had greater advantages in time and space complexity; in addition, the model can also predict unknown attacks.

I. Introduction

With the rapid development of the internet, the internet has become an indispensable part of people's work and life, and plays an increasingly important role in various fields. However, existing security measures and means cannot effectively defend against various network security threats. Therefore, how to effectively prevent network security threats has become a very important and urgent research topic. Among them, malicious encrypted traffic is a way of using encryption technology for attacks, and its strong concealment and high detection difficulty make it one of the main network attack methods.