An encrypted traffic analysis defense method and system based on an adversarial attack algorithm

Abstract

The invention discloses an encrypted traffic analysis and defense method and system based on an anti-attack algorithm, belonging to the field of network information security. Including: S1. The website traffic data of each session in the website traffic data sample is abstracted and compressed into a traffic burst sequence and used as a sample to obtain a sample set. Each element in the traffic burst sequence represents the continuous transmission or reception in the same direction. The number of data packets; S2. For each sample in the sample set, randomly select multiple samples with labels different from the current sample from the sample set as the target sample pool, and select the target sample closest to the current sample from the target sample pool; S3. Add disturbance to the current sample to make it approach the target sample, and judge whether the classification misjudgment probability reaches the threshold for the sample after filling the disturbance, if so, end, otherwise, go to step S2. The invention ensures the integrity and reliability of the network data, and at the same time ensures that the disturbance of the data is as small as possible to reduce the load consumption of the defense strategy.

Description

technical field [0001] The invention belongs to the field of network information security, and more specifically relates to a method and system for analyzing and defending encrypted traffic based on an anti-attack algorithm. Background technique [0002] With the increasing awareness of personal privacy protection of network users, the popularity of network data encryption methods is gradually increasing, and people pay more attention to personal privacy protection while pursuing efficient information acquisition efficiency. However, attackers can still use some data analysis methods to analyze encrypted or transmitted data using anonymous software. Current research shows that website fingerprinting technology is a common network attack method for encrypted traffic data, especially anonymous network data. How to formulate corresponding defense strategies for network attack forms of traffic data analysis is an urgent problem to be solved. In order not to affect the integrity...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the defects and improvement needs of the prior art encrypted traffic analysis and defense technology, which has large load overhead and poor operability, the present invention provides an encrypted traffic analysis and defense method and system based on an adversarial attack algorithm, and its purpose is to adopt deep The learning adversarial attack algorithm generates adversarial samples after perturbation and filling, so that the bandwidth load consumption generated by filling can be optimized, and the use of deep learning technology to analyze the characteristics of traffic data greatly reduces the complexity of the traffic monitoring algorithm in the defense process, making it possible Enhanced operability

Images