Unlock instant, AI-driven research and patent intelligence for your innovation.

Anti-confusion network intrusion detection method based on generative adversarial network

An intrusion detection and anti-aliasing technology, applied in biological neural network models, neural learning methods, instruments, etc., can solve problems such as unreasonable setting of discriminator training targets, lack of effectiveness of examples, difficulty in framework training, etc.

Active Publication Date: 2021-09-14
成都为辰信息科技有限公司 +1
View PDF14 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Firstly, the generated examples may lack validity, secondly, frame training may be difficult, and the training target setting of the discriminator may be unreasonable
In addition, for the confrontation examples generated by GAN, there is no effective evaluation method to verify its legitimacy and distribution rationality

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Anti-confusion network intrusion detection method based on generative adversarial network
  • Anti-confusion network intrusion detection method based on generative adversarial network
  • Anti-confusion network intrusion detection method based on generative adversarial network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0063] figure 2 This is the specific implementation flow chart of the anti-obfuscation network intrusion detection method based on the generative confrontation network of the present invention. like figure 2 As shown, the specific steps of the anti-obfuscation network intrusion detection method based on the generative adversarial network of the present invention include:

[0064] S201: Collection instance:

[0065] Collect several normal instances to form a normal instance set X_n, and collect several attack instances to form an attack instance set X_a. After determining the type of attack to be detected, each instance is divided into functional features and non-functional features. The functional features are: Those features that are closely related to the basic functions of the instance and if modified will destroy the effectiveness of the instance, these features are often strongly related to the type of attack to be detected, so such functional features cannot be modif...

example

[0111] Randomly select a set of instances from the instance set X to form the instance set collection of instances Each instance in is input to the intrusion detection model C for detection and input to the discriminator of the generative adversarial network for identification.

[0112] S506: Update the parameters of the intrusion detection model:

[0113] Use the following formula to calculate the objective function value of the intrusion detection model C, and update the parameters of the intrusion detection model C with the goal of maximizing the training objective function value:

[0114]

[0115] in, Represents a collection of instances the number of instances in Represents a collection of instances B an instance, Represents an instance 's true label, represents the intrusion detection model C for the instance detection score.

[0116] S507: Generator parameter update:

[0117] Use the following formula to calculate the objective function value o...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an anti-confusion network intrusion detection method based on a generative adversarial network, and the method comprises the steps: collecting a plurality of normal instances and attack instances, training the generative adversarial network through employing the attack instances, determining a target intrusion detection system which needs to improve the anti-confusion according to an actual condition, configuring an intrusion detection model for the target intrusion detection system and is trained, and then adopting an attack instance to perform joint training on a generator in the generative adversarial network and the intrusion detection model so as to realize cheating of the target intrusion detection system; and performing joint training on the generator and the intrusion detection model in the generative adversarial network again by adopting the normal instance and the attack instance so as to realize exceeding of a target intrusion detection system. Based on the attack instance generated by the generative adversarial network, the target intrusion detection system is simulated, cheated and surpassed by adopting one intrusion detection model, so that the anti-confusion performance of network intrusion is improved.

Description

technical field [0001] The invention belongs to the technical field of network intrusion detection, and more particularly relates to an anti-obfuscation network intrusion detection method based on a generative confrontation network. Background technique [0002] For network security, intrusion detection system is a crucial link, it is a tool configured on the router to detect network traffic. Intrusion detection system is divided into network intrusion detection system and host intrusion detection system. Network intrusion detection systems can identify malicious attacks from a large amount of network traffic. The host intrusion detection system can judge whether there are malicious behaviors and operations through its own related system call logs, and then detect threats to the system. In recent years, with the continuous development of computer technology, the computing performance of computers has been continuously improved, and the storage capacity has also been contin...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): H04L29/06G06N3/04G06N3/08G06K9/62
CPCH04L63/1441G06N3/08G06N3/045G06F18/24
Inventor 何俊鹏李允肖堃赵焕宇
Owner 成都为辰信息科技有限公司