Introduction: In the automotive industry, integrating connected and autonomous vehicles through the Internet of Vehicles and wireless communication has gained prominence. Secure communication and data privacy are the primary concerns in wireless communication. A weak privacy mechanism allows non-legitimate users to spoof for illegal or malicious activities or to manipulate the vehicular network. The robust Distributed Denial of Service (DDoS) detection and mitigation alone is required, as the risk of Distributed Denial of Service (DDoS) attacks is increasing in vehicular communication. The primary aim of this study is to design a secure and intelligent system for detecting and mitigating DDoS attacks in IoV. Specifically, the study aims to develop a hybrid detection model comprising Decision Tree and Random Forest classifiers to effectively analyze real-time network traffic. Another core goal is to ensure data privacy by implementing ElGamal encryption for all inter- vehicular communications. Additionally, the study proposes a trust-based algorithm that evaluates the reliability and trustworthiness of messages received by Roadside Units (RSUs) from registered vehicles. Methods: The proposed system analyses the vehicular network traffic to detect DDoS attacks using the hybrid approach combining Decision Tree and Random Forest algorithms. A trust-factor- based algorithm is designed to validate the event information the RSU receives from the registered vehicle. Depending on the trust factor and validation process, RSU checks for DDoS attacks and blocks vehicles that support the event. Privacy is ensured by encrypting all messages and information, including vehicle IDs and driver and passenger information, using the ElGamal encryption technique. Results: Experimental evaluation of the proposed hybrid model demonstrates superior performance in DDoS detection. The accuracy of the hybrid model is 99.81%, while those of logistic regression, kNN, and the Naive Bayes algorithm are 97.30%, 97.81%, and 97.10%, respectively. These results affirm the effectiveness of combining classification models with trust management and encryption for secure vehicular communication. Discussion: The proposed hybrid Decision Tree–Random Forest model achieves superior accuracy over existing methods while preserving data privacy through ElGamal encryption. Trust-factor validation strengthens detection but may lead to rare false blocks of legitimate vehicles. Future work includes restoration mechanisms, blockchain-based trust, and quantum computing for faster detection and post-quantum privacy. Conclusion: The proposed framework addresses critical vulnerabilities in current vehicular networks and offers a scalable, efficient, and patentable solution for intelligent transportation systems and vehicular cybersecurity.
Agarwal et al. (Fri,) studied this question.