Vehicle Reidentification in a Connected Vehicle Environment using Machine Learning Algorithms

Abstract

Deployment of Connected Vehicles will become available for most American cities in next to 20 years. The applications (e.g. mobility, safety, environment) are constantly receiving vehicle data. The current ID protection mechanism assumes a vehicle's ID changes every 5 minutes, the topic of re-matching vehicles is of interests in privacy protection and performance measure research. This paper explores the possibility of re-matching connected vehicles' ID using popular machine learning techniques, including Logistic Regression (LR), Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Linear and Nonlinear Support Vector Machine (SVM) and Nearest Neighbor algorithms. An experiment is conducted using a microscopic traffic simulation model through a Software-In-the-Loop technique. The best average mis-matching rate is 14%. To assess potential factors' effects on matching accuracy, a Poisson mixed regression model is analyzed under the Bayesian inference framework. Findings are: different matching algorithms vary in matching performance and the Linear SVM, the QDA and the LDA have the best accuracy results; traffic volume and market penetration rate have little impact on matching results; location and number of vehicles to be matched are considered significant. The results make the performance measure of future CV applications feasible and also suggest that more secure mechanisms are in need to protect the public.

Driver Back-Tracing Based on Automated Vehicle Identification Data

“…It is noted that the DBT problem is similar to the VPR problem to some extent. In the VPR problem, Miao et al applied conventional machine learning algorithms to re-match the changing IDs of CVs for privacy protection and performance measure ( 27 ). Yang et al applied a particle filter to completely reconstruct vehicle paths ( 2 ).…”

Privacy Implication of Location-Based Service: Multi-Class Stochastic User Equilibrium and Incentive Mechanism

“…While LBS has brought convenience to travelers, LBS platforms can easily obtain users’ location and trace information via positioning functions, and therefore pose risks to users’ privacy ( 1 ). Recently, Miao et al ( 2 ) discussed the likelihood of re-matching connected vehicles using structured connected-vehicle data. They compared seven machine learning methods and found the average success rate was about 86%.…”

Privacy Risk of Connected Vehicles in Relation to Vehicle Tracking when Transmitting Basic Safety Message Type 1 Data

“…Lu et al ( 14 ) recommend changing pseudonyms at social spots such as shopping malls or at intersections where multiple vehicles are sharing similar location and speed states to impede racking of the pseudonyms. Miao et al ( 15 ) evaluated seven popular machine learning algorithms to match the trajectory of the vehicles observed at one intersection to the trajectory of the vehicles at the next intersection under a policy of changing pseudonyms every five minutes. The result of their analysis of simulated data indicated the potential for matching trajectories after pseudonyms change.…”