Blockchain and AI-Enabled Trust Management Model for Internet of Vehicles
Mahalinoro Razafimanjato, Haishan Yang, Seri Park, Sunghyun Kim, Dongkyun Kim · 2025
The Internet of Vehicles (IoV) is a distributed network where connected vehicles and roadside units communicate seamlessly with one another and surrounding infrastructure. While this network facilitates enhanced inter-vehicular communication, its open and dynamically changing topology makes it vulnerable to the presence of malicious vehicles. These vehicles can transmit false and inaccurate messages, leading to severe, potentially life-threatening consequences for road users and compromising overall network security. Furthermore, the temporary and unreliable vehicle interaction can aggravate trust issues, hindering effective decision-making. However, existing trust management systems fall short of meeting the evolving demands of IoV, particularly in terms of accuracy, scalability, and real-time performance. We propose a novel trust management model for IoV to address these challenges. The model employs a Random Forest-based vehicle trust model to detect misbehavior and a data trust model using the Dempster-Shafer Theory to aggregate trust ratings from neighboring vehicles, comprehensively assessing event credibility. The final trust scores are securely stored on a permissioned blockchain using Hyperledger Fabric, ensuring the integrity of the trust scores while optimizing latency and throughput. The model's scalability and efficiency are validated through rigorous testing, significantly advancing IoV trust management.