CAN Bus Malicious Hardware Detection via Impedance Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for detecting malicious hardware on vehicle data communication networks, such as CAN bus, face interference issues when injecting signals, leading to distortions and potential errors, and lack effective methods to distinguish between authorized and unauthorized connections.
Innovation Solution
The system employs impedance measurement using AC or DC signals across a range of frequencies, including during Deep Sleep Mode, and incorporates statistical analysis with machine learning techniques to differentiate between legitimate and malicious units, minimizing interference with vehicle systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If signals are injected onto the CAN bus for detection purposes, then detection capability is improved, but signal interference and ECU distortion occur
Solution Approach 1:
The system performs impedance measurements periodically during Deep Sleep Mode intervals when no CAN communication is occurring, rather than continuously injecting signals. This periodic measurement approach enables detection functionality while avoiding continuous interference with ECU operations.
Solution Approach 2:
The system utilizes the existing Deep Sleep Mode intervals of the vehicle network as natural measurement opportunities, rather than requiring separate dedicated measurement time slots. This self-service approach leverages the vehicle's own operational cycles to perform security detections without additional interference.
2Measurement precision
If impedance measurements are performed during active communication, then detection accuracy is improved, but communication errors and system disruptions occur
Solution Approach 1:
The system schedules impedance measurements to occur periodically during Deep Sleep Mode when the CAN bus is inactive, ensuring that measurements are performed without disrupting active communication. This timing strategy maintains both detection accuracy and communication reliability.
Solution Approach 2:
The system performs impedance measurements in advance during Deep Sleep Mode intervals before normal communication resumes, allowing detection to occur proactively without interfering with subsequent communication operations.
3Productivity
If signal injection frequency is increased for faster detection, then detection speed is improved, but interference with vehicle systems increases
Solution Approach 1:
The system performs measurements during naturally occurring Deep Sleep Mode intervals, achieving detection without requiring high-frequency signal injection that would interfere with vehicle systems. The periodic nature of these intervals provides adequate detection speed while avoiding interference.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for reliable detection of malicious hardware without disrupting vehicle operations, providing accurate differentiation and alerting mechanisms for potential threats.
Implementation Method 1
an AC signal generating device (140) configured to generate AC electrical current
Implementation Method 2
an impedance measuring device (142) configured to measure the network bus impedance for each frequency
Data Source
AI summary
A system for detecting malicious hardware on a data communication network, such as a vehicle CAN bus, is provided. The system includes a teleprocessing device, an AC signal generating device, and an impedance measuring device. In a preliminary step, a set of impedance measurements of N reference AC signals is formed, and a threshold value is set. The signal generating device injects a set of N AC signals into the network bus and the bus impedance for each of the N frequencies is measured, where a set of impedance values of N RT-signals is formed. Then, each of the impedance values of the RT-signals and the impedance values of the respective reference AC signal are statistically compared, to thereby form a set of N comparison-results. Upon determining that any of the impedance values of the RT-signals is greater than the threshold, an alert is activated.