Blockchain-Based Trust Management Using Multi-Criteria Decision-Making Model for VANETs
Recent advancements in embedded sensing system, wireless communication technologies, big data, and artificial intelligence have fueled the development of Internet of Vehicles (IoV), where vehicles, road side unit (RSUs), and smart devices seamlessly interact with each other to enable the gathering and sharing of information on vehicles, roads, and their surrounds. As a fundamental component of IoV, vehicular networks (VANETs) are playing a critical role in processing, computing, and sharing travel-related information, which can help vehicles timely be aware of traffic situation and finally improve road safety and travel experience. However, due to the unique characteristics of vehicles, such as high mobility and sparse deployment making neighbor vehicles unacquainted and unknown to each other, VANETs are facing the challenge of evaluating the credibility of road safety messages. In this paper, we propose a blockchain-based trust management system using multi-criteria decision-making model, also referred to as TrustBlockMCDM, in VANETs. In the TrustBlockMCDM, each vehicle evaluates the credibility of received road safety message and generates the trust value of message originator. Due to the limited storage capacity, each vehicle periodically uploads the trust value to a nearby RSU. After receiving various trust values from vehicles, the RSU calculates the reputation value of message originator of road safety message using multi-criteria decision-making model, packs the reputation value into a block, and competes to add the block into blockchain. We evaluate the proposed TrustBlockMCDM approach through simulation experiments using OMNeT++ and compare its performance with prior blockchain-based decentralized trust management approach. The simulation results indicate that the proposed TrustBlockMCDM approach can not only improve fictitious message detection rate and malicious vehicle detection rate, but also can increase the number of dropped fictitious messages.