Network background traffic generation method based on generative adversarial network GAN

Abstract

The present invention proposes a network background traffic generation method based on a generative confrontation network GAN. The implementation steps are: 1) acquiring a training sample set; 2) constructing a generative confrontation network model library; 3) iteratively training the generative confrontation network model library; 4) Obtain the traffic data packet characteristics predicted by the trained generator network; 5) Network traffic generation results. The invention iteratively trains a model library composed of multiple generative adversarial networks of the same type as the network application through the training sample set containing the characteristics of the network traffic data packets of various network applications, thereby accelerating the convergence speed of the generative adversarial network model library. , On the premise of ensuring communication security, the efficiency of network background traffic generation is effectively improved.

Description

technical field [0001] The invention belongs to the technical field of network security, and relates to a method for generating network background traffic, in particular to a method for generating network background traffic based on a generative confrontation network GAN, which can be used to generate network background traffic. Background technique [0002] Communication nodes in the Internet need to interact with traffic data packets when using network applications to communicate. A network traffic sent by a communication node contains a set of packet sequences. in Indicates the ath that the communication node needs to send i traffic packets. [0003] Operators providing network application services need a large number of network traffic data packet samples for network security analysis and network stress testing, and network traffic generation technology is also constantly developing. The network traffic generation methods mainly include two kinds of network traffic ...

AI Technical Summary

Problems solved by technology

However, its shortcoming is that each traffic data packet sample in the training sample set must be filled to a fixed 1518 features after vectorization, and all different types of traffic data packets are used as a training sample set for a conditional generative adversarial network. In the process of obtaining the network background traffic generation results, iteratively training a conditional generative adversarial network through multiple different types of traffic data packets, the convergence speed is slow, resulting in low efficiency of network background traffic generation

Images