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

Protocol identification method based on hybrid model of residual network and recurrent neural network

A technology of cyclic neural network and protocol identification, applied in the field of network identification, can solve the problems of feature redundancy, poor model generalization ability, easy to miss important features, etc., and achieve the effect of improving accuracy

Pending Publication Date: 2022-02-11
ARMY ENG UNIV OF PLA
View PDF3 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This type of method can achieve better protocol recognition results, but manual design of features requires rich experience, and it is easy to miss important features, and there are often many redundant features in the feature set, and there are problems such as poor generalization ability of the model.

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
  • Protocol identification method based on hybrid model of residual network and recurrent neural network
  • Protocol identification method based on hybrid model of residual network and recurrent neural network
  • Protocol identification method based on hybrid model of residual network and recurrent neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0042] This embodiment introduces a protocol identification method based on a hybrid model of residual network and cyclic neural network, including:

[0043] Obtain network traffic data and form data packets with unknown protocol type;

[0044] Preprocessing the data packet and converting it into a one-dimensional vector;

[0045] The one-dimensional vector is input into the pre-built and trained protocol recognition model, the characteristics of the data packet are extracted, and the type of the application layer protocol corresponding to the data packet is judged by the model.

[0046] The protocol identification method based on the mixed model of residual network and cyclic neural network provided in this embodiment, its application process specifically involves the following steps:

[0047] (1) Data preprocessing: Data preprocessing is performed on the captured original network traffic, and application layer protocol data is extracted from the network traffic for subseque...

Embodiment 2

[0069] This embodiment provides a protocol identification device based on a hybrid model of residual network and cyclic neural network, including:

[0070] An acquisition unit, configured to acquire network traffic data and form data packets with unknown protocol types;

[0071] A preprocessing unit, configured to preprocess the data packet and convert it into a one-dimensional vector;

[0072] The judging module is used to input the one-dimensional vector into the pre-built and trained protocol recognition model, extract the characteristics of the data packet, and judge the application layer protocol type corresponding to the data packet through the model.

Embodiment 3

[0074] This embodiment provides a protocol identification device based on a mixed model of residual network and cyclic neural network, including a processor and a storage medium;

[0075] The storage medium is used to store instructions;

[0076] The processor is configured to operate in accordance with the instructions to perform the steps of any of the following methods:

[0077] Obtain network traffic data and form data packets with unknown protocol type;

[0078] Preprocessing the data packet and converting it into a one-dimensional vector;

[0079] The one-dimensional vector is input into the pre-built and trained protocol recognition model, the characteristics of the data packet are extracted, and the type of the application layer protocol corresponding to the data packet is judged by the model.

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 a protocol identification method based on a hybrid model of a residual network and a recurrent neural network, and belongs to the technical field of networks, and the method comprises the steps: obtaining network flow data, and forming a data packet with an unknown protocol type; preprocessing the data packet, and converting the data packet into a one-dimensional vector; inputting the one-dimensional vector into a pre-constructed and trained protocol identification model, extracting features of the data packet, and judging an application layer protocol category corresponding to the data packet through the model. According to the method, the spatial features of the protocol data are extracted by using the one-dimensional pre-activation residual network, then the time features of the protocol data are extracted by using the bidirectional gating recurrent neural network, and finally the key features of the protocol are further extracted by using the attention mechanism to implement protocol classification, so the accuracy of network protocol identification is effectively improved.

Description

technical field [0001] The invention relates to a protocol recognition method based on a mixed model of a residual network and a cyclic neural network, and belongs to the technical field of network recognition. Background technique [0002] The core purpose of protocol identification is to identify the application layer protocol to which the network communication flow belongs, which is the core technology of network security, network management and network service quality assurance. Existing protocol identification methods can be mainly divided into four categories: port-based identification methods, deep packet inspection-based identification methods, traditional machine learning-based identification methods, and deep learning-based identification methods. [0003] The port-based protocol identification method mainly identifies the application layer protocol based on the port number. For example, the application layer protocol based on the TCP protocol for transmission and ...

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): H04L43/18G06N3/04G06N3/08
CPCH04L43/18G06N3/08G06N3/045
Inventor 洪征吴吉胜林培鸿张沈梅马甜甜
Owner ARMY ENG UNIV OF PLA