In-Vehicle Gateway Autoencoder for CAN Bus Intrusion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern vehicles' Electronic Control Units (ECUs) are vulnerable to malicious network packets via V2X, Bluetooth, and USB connections, which can compromise vehicle operations and passenger safety due to the lack of a mechanism to prevent such threats from reaching the CAN bus.
Innovation Solution
An in-vehicle network security system that uses machine learning models, specifically autoencoders, to generate input vectors from messages received on the CAN bus, determine abnormal patterns, and prevent malicious messages from entering the bus by comparing reconstruction losses against a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If V2X, Bluetooth, and USB connections are enabled for vehicle networking services, then vehicle functionality and connectivity are improved, but security vulnerability increases due to potential malicious network packets
Solution Approach 1:
The gateway acts as an intermediary security mechanism between external network connections (V2X, Bluetooth, USB) and the internal CAN bus. It receives messages from external interfaces, analyzes them using machine learning models, and selectively forwards only safe messages to the CAN bus, preventing malicious packets from reaching ECUs while maintaining legitimate communication functionality
2Measurement precision
If machine learning models are used to detect abnormal messages, then detection accuracy is improved, but system complexity increases
Solution Approach 1:
The system uses unsupervised learning where the machine learning model automatically learns normal message patterns from CAN bus traffic without requiring manual labeling or complex feature engineering. The autoencoder architecture self-adjusts to capture temporal dependencies and anomaly patterns, reducing the need for complex manual configuration while maintaining high detection accuracy
Solution Approach 2:
The system transforms raw message data into feature vectors with specific dimensionalities suitable for machine learning processing. By changing the representation parameters of messages into a standardized vector format that can be processed by the autoencoder, the system simplifies the detection mechanism while improving accuracy through learned patterns
Data Source
AI summary
In one example, a method comprises: receiving, at a gateway of a Controller Area Network (CAN) bus on a vehicle and via at least one of a wireless interface or a wired interface of the vehicle, a plurality of messages targeted at electronic control units (ECU) on the CAN bus; generating one or more input vectors based on the plurality of messages; generating, using one or more machine learning models, an output vector based on each of the one or more input vectors, each input vector having the same number of elements as the corresponding output vector; generating one or more comparison results between each of the one or more input vectors and the corresponding output vector; and based on the one or more comparison results, performing one of: allowing the plurality of messages to enter the CAN bus or preventing the plurality of messages from entering the CAN bus.