Hybrid Machine Learning and Trust-Based Routing for Enhanced Misbehavior Detection in VANETs

Abstract

Wireless Vehicular Networks (VANETs) are a key component of Intelligent Transportation Systems (ITS), but they face significant security challenges due to their dynamic nature and reliance on wireless communication. To address these challenges, this project proposes a hybrid security framework that combines machine learning and trust-based routing to enhance the detection of malicious behaviors and improve routing efficiency within the network. The proposed framework consists of two main components: a machine learning approach, which employs the Random Forest algorithm in a Python environment to detect manipulation in vehicles’ physical coordinates by analyzing features like position drift and RSSI; and a network-based component, implemented using OMNeT++ and the Veins framework, which integrates a trust- based system to evaluate node reliability and enhance routing by modifying the traditional GPSR protocol. The results of both components demonstrated the effectiveness of this framework in detecting abnormal behaviors and improving communication quality, thereby enhancing the security of VANETs in complex environment