Vehicle Data Analysis for Cyberattack Detection via External Cross-Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing security technologies for in-vehicle networks struggle to accurately detect sophisticated cyberattacks, particularly those that mimic normal data, and fail to identify and isolate equipment transmitting illegal data, leading to potential damage expansion through extended data flow paths.
Innovation Solution
A data analysis apparatus that obtains vehicle data and external data from nearby vehicles and traffic infrastructure systems to determine inconsistencies, enabling more accurate anomaly detection and preventing damage propagation by transmitting alerts and instructions to relevant systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only in-vehicle CAN data is analyzed for security detection, then the analysis system remains simple and fast, but sophisticated attacks using spoofing tactics cannot be accurately sensed
Solution Approach 1:
The patent merges in-vehicle CAN data analysis with external data (traffic information from other vehicles and roadside units) to create a comprehensive security detection system. This combination enables cross-validation of data authenticity, allowing the system to detect sophisticated spoofing attacks that would be indistinguishable from legitimate data when analyzed in isolation.
Solution Approach 2:
The patent introduces external infrastructure (roadside units and other vehicles) as intermediaries to verify the authenticity of in-vehicle data. These intermediaries provide independent data sources that can confirm or refute claims made by the vehicle's own sensors, enabling detection of spoofed information without requiring complex internal analysis of the vehicle's own systems.
2Measurement precision
If external data from other vehicles and infrastructure is incorporated for verification, then attack detection accuracy improves, but data processing time and system complexity increase
Solution Approach 1:
The patent performs preliminary filtering and validation of external data before full analysis. The system pre-establishes trust relationships with roadside units and other vehicles, and pre-processes incoming data streams to identify obviously inconsistent or malformed data early in the processing pipeline, reducing the time required for comprehensive verification.
Solution Approach 2:
The patent implements a tiered verification approach where not all external data requires full validation. The system performs partial verification on high-volume data streams and reserves comprehensive verification for data points that trigger anomaly detection thresholds, balancing processing time with detection accuracy by applying verification effort proportionally to suspected risk levels.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A data analysis apparatus includes: a vehicle data obtainer (210) that obtains vehicle data indicating a vehicle status of a first vehicle; a external data obtainer (210) that obtains external data indicating external circumstances of the first vehicle from at least one of a second vehicle, which is in an area where the first vehicle is running and is different from the first vehicle, and a traffic infrastructure system; and a determiner (230) that performing a first determination as to whether there is an inconsistency between the vehicle status of the first vehicle indicated and the external circumstances, and outputs a result of the first determination regarding the first vehicle.