Secured Traffic Monitoring in VANET

TL;DR

This paper presents a privacy-preserving, edge cloud-based model for secure traffic monitoring in VANETs that effectively authenticates and filters malicious information, ensuring accurate traffic data even under attack scenarios.

Contribution

The paper introduces a novel heuristic-based authentication model leveraging vehicular data, validated through simulations, improving security and accuracy in VANET traffic monitoring.

Findings

Effective filtering of malicious vehicles in simulations

Accurate traffic information maintained under attack scenarios

Outperforms peer-based and reputation systems in robustness

Abstract

Vehicular Ad hoc Networks (VANETs) facilitate vehicles to wirelessly communicate with neighboring vehicles as well as with roadside units (RSUs). However, the existence of inaccurate information within the network can cause traffic aberrations and also disrupt the normal functioning of any traffic monitoring system. Thus, determining the credibility of broadcast messages originating from the region of interest (ROI) is crucial under a malicious environment. Additionally, a breach of privacy involving a vehicle's private information, such as location and velocity, can lead to severe consequences like unauthorized tracking and masquerading attack. Thus, we propose an edge cloud based privacy-preserving secured decision making model that employs a heuristic based on vehicular data such as GPS location and velocity to authenticate traffic-related information from the ROI under different traffic scenarios such as congestion. The effectiveness of the proposed model has been validated using VENTOS, SUMO, and Omnet++ simulators, and also by using a simulated cloud environment. We compare our proposed model to the existing peer-based authentication model, the majority voting model, and the reputation-based system under different attack scenarios. We show that our model is capable of filtering malicious vehicles effectively and provide accurate traffic information under the presence of at least one non-malicious vehicle within the ROI.