Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Vehicle-mounted CAN intrusion detection method based on sliding window and CENN

A sliding window, intrusion detection technology, applied in neural learning methods, biological neural network models, protection of internal/peripheral computer components, etc., can solve problems such as high false alarm rate and lack of feature preprocessing methods

Active Publication Date: 2020-11-13
CHONGQING UNIV OF POSTS & TELECOMM
View PDF10 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For example: convolutional neural network (CNN), LSTM, etc. are applied to the detection scene, but only the relatively easy-to-handle CAN ID field is used in the feature, lacking a systematic feature preprocessing method, and the results obtained by these networks have high false positive rate

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
  • Vehicle-mounted CAN intrusion detection method based on sliding window and CENN
  • Vehicle-mounted CAN intrusion detection method based on sliding window and CENN
  • Vehicle-mounted CAN intrusion detection method based on sliding window and CENN

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0038] This embodiment provides an implementation of setting a square window according to the obtained feature size and obtaining the features of the original data according to the sliding window. Perform data conversion and convert the data into binary data, where CAN ID occupies 29 bits, and less than 29 bits will be filled with 0. For 64-bit CAN Data, it is converted in the same way as CAN ID, that is, CAN Data occupies 64 bits, and less than 64 bits will be filled with 0.

[0039] The obtained binary information is expanded, that is, the sum of CAN ID and CAN Data converted binary dimensions is used as the expanded feature size.

[0040] Divide all data records into square windows, where each window is a 93x93 two-dimensional matrix, and set the processing window size to a 93x93 square matrix, where the kth window is expressed as:

[0041]

[0042] Among them, SW k Indicates the kth window, x {i，j} The binary representation of the raw data, its value is 0 or 1.

[0...

Embodiment 2

[0047] On the basis of Embodiment 1, this embodiment provides a method for predicting data based on a convolutional encoder neural network.

[0048] In this embodiment, the encoding part of the convolutional neural network includes an encoder and an Inception-Resnet structure, and the encoder includes two convolutional layers and two pooling layers for reducing the sliding window from 93×93×1 dimension to To 6×6×64 dimension; Inception-Resnet structure, including 3 parallel convolutional layers, used to reduce the data from 6×6×64 dimension to 3×3×64 dimension.

[0049] The convolutional encoder network (CEN) proposed by the present invention is composed of the basic structure of CNN and the partial structure of autoencoder (AutoEncoder). AutoEncoder has the ability to capture nonlinear dependencies and reduce feature size. Such as image 3 As shown, AutoEncoder is divided into encoder and decoder. The implementation of encoder and decoder is described as follows:

[0050] ...

Embodiment 3

[0054] In this embodiment, on the basis of Embodiment 2, the prediction result obtained by the convolutional neural network is further described.

[0055] Studies have shown that dropout (half of the hidden node values ​​are set to 0) is a better choice for training deep neural networks, because it can avoid overfitting and improve the accuracy of the underlying neural network. Therefore, this embodiment first expands the network (corresponding to Figure 5 Average pooling in the previous operation), then use Dropout to prevent the model from overfitting, and output the result to the fully connected layer, and use SoftMax to obtain the training result.

[0056] exist Figure 5 Among them, the Encoder Structure part is the encoder, which reduces the dimensionality of the input 93×93×1 dimensional data to 47×47×32 through the convolutional layer (Covn1), and becomes 24 through a maximum pooling layer (Max-pooling). ×24×32 dimensions, then converted to 12×12×64 dimensions throu...

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 relates to the technical field of vehicle-mounted network intrusion detection, in particular to a vehicle-mounted controller local area network intrusion detection method based on a sliding window and a convolutional encoder neural network. The method comprises the steps: carrying out the data cleaning of received original data, and selecting a CAN ID and a CAN Data as features; converting the characteristics of each dimension of the CAN ID and the CAN Data into binary data; calculating a feature size according to the digit of the converted binary data; setting a square window according to the obtained feature size, and performing sliding window processing on the data by using the window to obtain processed features; inputting the processed features into a convolutional encoder neural network, wherein the convolutional encoder neural network obtains a detection result according to the input data. According to the invention, the abnormal flow and the normal flow can be discriminated with a low false alarm rate and a high detection rate, and the purpose of CAN network intrusion detection is achieved.

Description

technical field [0001] The invention relates to the technical field of intrusion detection, in particular to an intrusion detection method for a vehicle-mounted Controller Area Network (CAN) based on a sliding window and a convolutional encoder neural network (Convolutional Encoder Neural Networks, CENN). Background technique [0002] Dozens or even hundreds of various electronic control units (ECUs) are usually integrated in modern vehicles. These ECUs, intelligent sensors and in-vehicle networks form the basis of modern smart cars. In-vehicle network There are five widely used in-vehicle networks according to the information volume, response speed, reliability, application, and other requirements of the system. Table 1 shows the advantages and disadvantages of these five in-vehicle networks, namely Local Interconnect Network (LIN), Controller Area Network (CAN), FlexRay, Media Oriented System Transport (MOST) and Ethernet. [0003] Table 1 Comparison of in-vehicle network...

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): G06F21/71G06F21/85G06N3/04G06N3/08
CPCG06F21/71G06F21/85G06N3/049G06N3/08G06N3/047G06N3/045
Inventor 崔晓通程克非张兴张亮
Owner CHONGQING UNIV OF POSTS & TELECOMM
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com