Vehicle Bus Message Anomaly Detection With Self-Learning Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle bus systems lack efficient methods to detect deviations from typical communication, making them susceptible to manipulation without detection.
Innovation Solution
A control unit equipped with a learning method, such as a neural network, is trained to recognize typical bus communication patterns and detect deviations, providing an alarm when manipulation is detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a control unit uses a learning method to detect manipulation of bus messages, then detection accuracy is improved, but device complexity increases
Solution Approach 1:
The control unit performs self-learning by automatically analyzing received bus messages and establishing communication patterns without external intervention. The system trains its own learning method using actual message traffic, enabling it to detect manipulations while reducing the need for complex pre-programmed detection rules
Solution Approach 2:
The patent replaces traditional rule-based or threshold-based detection mechanisms with a learning method (such as neural networks or machine learning algorithms). This substitution enables the system to adaptively learn communication patterns and detect anomalies, improving detection accuracy while the learning algorithm handles the complexity internally
2Adaptability or versatility
If the learning method is trained using received messages, then adaptability to control unit changes is improved, but loss of time for training occurs
Solution Approach 1:
The control unit performs training operations in advance by continuously learning from incoming bus messages during normal operation. The learning method accumulates training data and adjusts its parameters proactively, so that when control unit changes occur (such as controller replacements or software updates), the system is already adapted or can adapt quickly without significant operational disruption
Solution Approach 2:
The learning method operates continuously in the background during normal bus communication, constantly refining its understanding of communication patterns. This continuous training ensures the system remains adaptive to changes while minimizing interruption to the primary function of message processing, effectively overlapping training with operational time
Data Source
AI summary
A method of using a controller of a vehicle to detect manipulation of a message of a bus system of the vehicle is disclosed herein. The method includes receiving, at the controller, the message from the bus system of the vehicle, and ascertaining a state of a learning method of the controller based on a vehicle state. When the state of the learning method indicates that the learning method is not trained: the method further includes training the learning method using the received message. When the state of the learning method indicates that the learning method is trained, the method further includes using the trained learning method to detect manipulation of the message. The method also includes providing an alarm message from the controller to a server outside of the vehicle when the trained learning method has detected manipulation of the message.