A Computational Model for Reputation and Ensemble-Based Learning Model for Prediction of Trustworthiness in Vehicular Ad Hoc Network

Alharthi, Abdullah and Ni, Qiang and Jiang, Richard and Khan, Mohammad Ayoub (2023) A Computational Model for Reputation and Ensemble-Based Learning Model for Prediction of Trustworthiness in Vehicular Ad Hoc Network. IEEE Internet of Things Journal, 10 (20): 20. pp. 18248-18258. ISSN 2327-4662

Text (IoTJ_Author final manuscript)
IoTJ_Author_final_manuscript.pdf - Accepted Version
Available under License Creative Commons Attribution-NonCommercial.
Download (1MB)

Abstract

Vehicular ad hoc networks (VANETs) are a special kind of wireless communication network that facilitates vehicle-to-vehicle(V2V) and vehicle-to-infrastructure(V2I) communication. This technology exhibits the potential to enhance the safety of roads, efficiency of traffic, and comfort of passengers. However, this can lead to potential safety hazards and security risks, especially in autonomous vehicles that rely heavily on communication with other vehicles and infrastructure. Trust, the precision of data, and the reliability of data transmitted through the communication channel are the major problems in VANET. Cryptography-based solutions have been successful in ensuring the security of data transmission. However, there is still a need for further research to address the issue of fraudulent messages being sent from a legitimate sender. As a result, in this study, we have proposed a methodology for computing vehicles reputation and subsequently predicting the trustworthiness of vehicles in networks. The blockchain records the most recent assessment of the vehicle’s credibility. This will allow for greater transparency and trust in the vehicle’s history, as well as reduce the risk of fraud or tampering with the information. The trustworthiness of a vehicle is confirmed not just by the credibility, but also by its network behavior as observed during data transfer. To classify the trust, an ensemble learning model is used. In depth tests are run on the dataset to assess the effectiveness of the proposed ensemble learning with feature selection technique. The findings show that the proposed ensemble learning technique achieves a 99.98% accuracy rate, which is notably superior to the accuracy rates of the baseline models.