A Deep Packet Inspection Method and System Based on Generative Adversarial Networks

Abstract

The invention discloses a deep packet detection system based on generating confrontation networks: the current policy rule base is divided into a first set A and a second set B; the generation network G in DPI-L-GAN accepts the first set A and the second set B; The traffic characteristics F mined from the actual network traffic are taken as input, and the generation network G of DPI-E-GAN accepts the first set A and the traffic characteristics F mined from the actual network traffic and the simulated network traffic S generated by the simulated traffic generator as input, DPI The generation network G of ‑L‑GAN and DPI‑E‑GAN is based on the corresponding input to generate a new rule strategy Rn; Rn and the second set B are input into the decision network D of DPI‑L‑GAN and DPI‑E‑GAN to determine the generation The logical validity of the rules and strategies; use S to evaluate Rn to determine the functional effectiveness of Rn; add effective Rn to the policy rule base, and check and update the policy rule base at the same time. The deep packet inspection system of the present invention is more intelligent and more scalable. The invention also provides a corresponding deep packet detection method based on generating confrontation networks.

Description

A Generative Adversarial Network-Based Deep Packet Detection Method and System technical field The present invention belongs to the technical field of data communication networks, and more particularly, relates to a deep neural network based on generative confrontation network. A packet detection method and system. Background technique [0002] The continuous development of the Internet has made the network more and more complex, and the operation and maintenance management of the network has become more difficult. Network security issues are becoming increasingly prominent. All kinds of network attack traffic have brought harm to the network, and all kinds of network garbage flooded the network network, which consumes network resources and has all the harm and no benefit. These require corresponding technologies and tools to identify and suppress. at the same time, The variety of traffic in the Internet requires technology and tools to accurately perceive and classi...

AI Technical Summary

Problems solved by technology

Although the focus of deep packet inspection technology is to summarize application characteristics and identify traffic based on application characteristics, traditional deep packet inspection technologies are also unable to cope with frequently mutating application characteristics.

Images