Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

GPS Spoofing Attack Detection in Autonomous Vehicles Using Adaptive DBSCAN

Ahmad Mohammadi, Reza Ahmari, Vahid Hemmati, Frederick Owusu-Ambrose, Mahmoud Nabil Mahmoud, Parham Kebria, Abdollah Homaifar, Mehrdad Saif

TL;DR

The paper addresses GPS spoofing in autonomous vehicles, focusing on both abrupt and gradual multi-step attacks. It introduces a two-stage framework: a displacement-predicting neural network for detecting large-magnitude spoofing and an adaptive DBSCAN with a dynamic ε for sub-threshold, small biases, all calibrated from clean data. The approach demonstrates near-perfect to excellent detection across several spoofing scenarios on HDD data, underscoring its potential for real-time AV security. By coupling data-driven displacement estimates with adaptive clustering, the method enhances robustness against GPS spoofing and improves safety in autonomous navigation.

Abstract

As autonomous vehicles become an essential component of modern transportation, they are increasingly vulnerable to threats such as GPS spoofing attacks. This study presents an adaptive detection approach utilizing a dynamically tuned Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, designed to adjust the detection threshold (ε) in real-time. The threshold is updated based on the recursive mean and standard deviation of displacement errors between GPS and in-vehicle sensors data, but only at instances classified as non-anomalous. Furthermore, an initial threshold, determined from 120,000 clean data samples, ensures the capability to identify even subtle and gradual GPS spoofing attempts from the beginning. To assess the performance of the proposed method, five different subsets from the real-world Honda Research Institute Driving Dataset (HDD) are selected to simulate both large and small magnitude GPS spoofing attacks. The modified algorithm effectively identifies turn-by-turn, stop, overshoot, and multiple small biased spoofing attacks, achieving detection accuracies of 98.621%, 99.960.1%, 99.880.1%, and 98.380.1%, respectively. This work provides a substantial advancement in enhancing the security and safety of AVs against GPS spoofing threats.

GPS Spoofing Attack Detection in Autonomous Vehicles Using Adaptive DBSCAN

TL;DR

The paper addresses GPS spoofing in autonomous vehicles, focusing on both abrupt and gradual multi-step attacks. It introduces a two-stage framework: a displacement-predicting neural network for detecting large-magnitude spoofing and an adaptive DBSCAN with a dynamic ε for sub-threshold, small biases, all calibrated from clean data. The approach demonstrates near-perfect to excellent detection across several spoofing scenarios on HDD data, underscoring its potential for real-time AV security. By coupling data-driven displacement estimates with adaptive clustering, the method enhances robustness against GPS spoofing and improves safety in autonomous navigation.

Abstract

As autonomous vehicles become an essential component of modern transportation, they are increasingly vulnerable to threats such as GPS spoofing attacks. This study presents an adaptive detection approach utilizing a dynamically tuned Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, designed to adjust the detection threshold (ε) in real-time. The threshold is updated based on the recursive mean and standard deviation of displacement errors between GPS and in-vehicle sensors data, but only at instances classified as non-anomalous. Furthermore, an initial threshold, determined from 120,000 clean data samples, ensures the capability to identify even subtle and gradual GPS spoofing attempts from the beginning. To assess the performance of the proposed method, five different subsets from the real-world Honda Research Institute Driving Dataset (HDD) are selected to simulate both large and small magnitude GPS spoofing attacks. The modified algorithm effectively identifies turn-by-turn, stop, overshoot, and multiple small biased spoofing attacks, achieving detection accuracies of 98.621%, 99.960.1%, 99.880.1%, and 98.380.1%, respectively. This work provides a substantial advancement in enhancing the security and safety of AVs against GPS spoofing threats.

Paper Structure

This paper contains 11 sections, 2 equations, 3 figures, 2 tables, 1 algorithm.

Table of Contents

  1. INTRODUCTION
  2. RELATED WORKS
  3. PROPOSED METHODOLOGY
  4. Detection of Single Large Magnitude Attacks
  5. Detection of Multiple Small Biased Attacks
  6. EVALUATION AND RESULTS
  7. ANN Model Development and the Detection of Single Large Magnitude Attacks
  8. Model Development
  9. Attack Detection
  10. Detection of Multiple Small Biased Attacks with Magnitudes Smaller than System Threshold
  11. CONCLUSIONS AND FUTURE WORK

Figures (3)

  • Figure 1: Representation of a single attack with a magnitude of 10 m at t=3 second and a series of multiple small biased attacks applied in 10 steps (1 m in each step) to achieve the same attack magnitude.
  • Figure 2: The proposed framework for detecting GPS spoofing attacks: multiple small biased attacks are detected with adaptive DBSCAN by using recursive mean and STD calculation from the uncertainty in clean datasets.
  • Figure 3: (a) The proposed neural network receives ten consecutive samples of speed, $\cos(\psi)$, $\sin(\psi)$, and time steps as input, while the GPS-perceived displacement—sampled every tenth data point—serves as the target output. (b) The training and test loss curves, along with performance metrics, demonstrate a steady decline in error over 200 epochs, indicating a smooth learning process.