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

HTM-based vehicle-mounted CAN network anomaly detection method and system

A technology for network anomalies and detection methods, applied in bus networks, transmission systems, digital transmission systems, etc., can solve the problems of poor applicability and sensitivity, failure to ensure the security of the vehicle CAN network, and the lightness and sensitivity of the detection model Insufficient and other problems, to achieve the effect of improving sensitivity and accuracy

Inactive Publication Date: 2022-03-25
SHANGHAI JIAO TONG UNIV +1
View PDF6 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the applicability and sensitivity of this patent are poor, and the security of the vehicle CAN network cannot be well guaranteed.
[0006] The Chinese patent with the application number 202110143987.2 in the prior art discloses "a low-latency and safe vehicle-mounted intrusion detection method based on deep learning", which uses convolutional neural network and attention mechanism to extract the depth of CAN messages The detection model constructed by this method is complex, and the detection accuracy rate is high only when the abnormal packets in the experimental data account for 20%, and there are deficiencies in the lightness and sensitivity of the detection 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
  • HTM-based vehicle-mounted CAN network anomaly detection method and system
  • HTM-based vehicle-mounted CAN network anomaly detection method and system
  • HTM-based vehicle-mounted CAN network anomaly detection method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0038] The invention detects the abnormal message in the vehicle CAN network based on the HTM model, so as to detect the attack on the vehicle system. Specifically, obtain the ID sequence of the CAN message during the driving process of the vehicle, construct the HTM model, train the HTM model after encoding the ID sequence of the CAN message generated under the normal driving of the vehicle, and use the trained HTM model to detect whether the vehicle is actually driving. Whether there is abnormal behavior, and then determine whether there is an attack. The HTM algorithm-based vehicle CAN network anomaly detection method of the present invention can detect attacks keenly as long as there are few abnormal changes in the vehicle CAN network data flow, and usually the proportion of abnormal messages is less than 5%.

[0039] The present invention provides a kind of CAN network anomaly detection method based on HTM algorithm, comprising:

[0040] Step 1, collecting data, wherein ...

Embodiment 2

[0064] Embodiment 2 is a preferred example of Embodiment 1.

[0065] Such as figure 1 , is an overall structural diagram of an HTM-based vehicle CAN network anomaly detection system provided by an embodiment of the present invention, including:

[0066] The data collection unit is used to collect CAN messages generated during vehicle driving, and extract ID information in the messages to form an ID sequence.

[0067] The data preprocessing unit is used to convert the obtained message ID sequence data into a binary vector of uniform format and length, which is used as the input of the anomaly detection unit.

[0068] The anomaly detection unit builds and trains the HTM model, uses the model to predict the ID input data, calculates the abnormal value, and then evaluates the security of the CAN network.

[0069] As another optional mode in the examples of the present invention, the abnormality detection unit includes:

[0070] The sparse coding module is used to perform sparse...

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 provides a vehicle-mounted CAN network anomaly detection method and system based on an HTM. The method comprises the steps that various CAN messages generated by a vehicle-mounted CAN network in the vehicle driving process are acquired; the obtained message ID sequence is preprocessed, and sparse binary vectors are generated through sparsification; learning the input sparse binary vector by using an HTM algorithm, generating a predicted value of the next moment according to the input data of the current moment, comparing an actually input message with a predicted result, comparing the difference between the actually input message and the predicted result, and further converting the difference into an abnormal value; setting an abnormal threshold value, judging whether the abnormal value is greater than the abnormal threshold value, and deciding whether to declare abnormal. The vehicle-mounted CAN network flow detection method can discover subtle flow changes of the vehicle-mounted CAN network so as to discover malicious attack behaviors in time, has high sensitivity, consumes less vehicle resources, and is convenient for practical application.

Description

technical field [0001] The invention relates to the technical field of in-vehicle CAN network security, in particular to an HTM-based method and system for detecting anomalies in a vehicle CAN network. Background technique [0002] With the increasing demand for intelligent interconnection of automobiles, more and more automobile manufacturers are gradually launching automobiles with Internet connection functions, and the interaction between automobiles and the outside world is becoming more and more extensive. In the intelligent network environment, the vehicle network structure and functions are becoming more and more complex, the attack surface of the vehicle is becoming wider and wider, and the security threats it faces are becoming more and more serious. [0003] At present, there are mainly two security strategies based on prevention and detection to ensure the security of vehicle network. The prevention-based security strategy uses security mechanisms such as encrypt...

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
Patent Type & Authority Applications(China)
IPC IPC(8): H04L9/40H04L12/40H04L67/12
CPCH04L63/1425H04L67/12H04L12/40H04L2012/40215
Inventor 姚立红阚晓鹏蒋兴浩訾小超张月国
Owner SHANGHAI JIAO TONG UNIV