Temporal Consistency Checks to Detect LiDAR Spoofing Attacks on Autonomous Vehicle Perception
This addresses a critical security vulnerability for autonomous vehicles, offering a novel detection approach to prevent hazardous driving decisions from spoofing attacks.
The paper tackles the problem of LiDAR spoofing attacks on autonomous vehicles by proposing a method that uses motion as a physical invariant to detect fake objects, achieving over 98% attack detection rate and 91% recall for spoofed vehicles with real-time performance at 41Hz.
LiDAR sensors are used widely in Autonomous Vehicles for better perceiving the environment which enables safer driving decisions. Recent work has demonstrated serious LiDAR spoofing attacks with alarming consequences. In particular, model-level LiDAR spoofing attacks aim to inject fake depth measurements to elicit ghost objects that are erroneously detected by 3D Object Detectors, resulting in hazardous driving decisions. In this work, we explore the use of motion as a physical invariant of genuine objects for detecting such attacks. Based on this, we propose a general methodology, 3D Temporal Consistency Check (3D-TC2), which leverages spatio-temporal information from motion prediction to verify objects detected by 3D Object Detectors. Our preliminary design and implementation of a 3D-TC2 prototype demonstrates very promising performance, providing more than 98% attack detection rate with a recall of 91% for detecting spoofed Vehicle (Car) objects, and is able to achieve real-time detection at 41Hz