Autonomous Navigation Attack Detection Using VIO Cross-Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cyber-attacks targeting sensor data in autonomous vehicles corrupt navigation systems, leading to unreliable control algorithms and navigation data.
Innovation Solution
Implementing Replay-Attack Detection Using Pose Validation and GPS Spoofing Detection techniques, augmented with root-of-trust hardware, to cross-validate sensor data using inertial measurement units, neural networks, and trusted on-board hardware oscillators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If sensor data is used for autonomous navigation, then navigation functionality is enabled, but the system becomes vulnerable to cyber-attacks and sensor data corruption
Solution Approach 1:
The system implements feedback mechanisms by continuously monitoring sensor data through multiple independent sensors (IMU, barometer, camera) and comparing their readings against expected physical relationships. When discrepancies are detected, the system feeds this information back to detect and respond to potential attacks, thereby maintaining reliability while preserving navigation functionality.
Solution Approach 2:
The patent introduces intermediary components including a barometer as a mediator between GPS and ground truth determination, and uses camera imagery as an intermediary to validate GPS position data. These intermediaries provide additional verification layers that prevent direct exploitation of vulnerable sensors while maintaining overall system functionality.
2Measurement precision
If multiple sensors are used for cross-validation, then detection accuracy improves, but device complexity increases
Solution Approach 1:
The patent applies multi-functionality by using the camera system for dual purposes: primary navigation/visual odometry and secondary security function for GPS validation. The IMU and barometer serve both navigation and attack detection functions. This universal usage of sensors reduces the need for dedicated additional hardware while improving detection accuracy.
Solution Approach 2:
The system employs self-service principles by using its existing navigation sensors (IMU, barometer, camera) to perform security validation functions. Rather than requiring entirely separate dedicated security hardware, the navigation system's own components validate each other's data, reducing overall system complexity while maintaining detection accuracy.
Data Source
AI summary
Autonomous navigation cyber-attack detection and/or avoidance techniques include visual and inertial odometry (VIO) algorithms to provide a root-of-trust during navigation, VIO algorithms that cross-validate navigation parameters using IMU and visual data, and hardware-dependent attack survival mechanisms that support autonomous systems during an attack.