A Feedback-Driven Lightweight Reputation Scheme for IoV

Most applications of Internet of Vehicles (IoVs) rely on collaborations between nodes. It is therefore of vital importance in building up the implicit trust between these nodes, the false information flow in-between these nodes poses the challenging trust issue in rapidly moving IoV nodes. To resolve this issue, a number of mechanisms have been proposed in the literature for the detection of false information and establishment of trust in IoVs, most of which employ reputation scores as one of the important factors. However, it is critical to have a robust and consistent scheme that is suitable to aggregate a reputation score for each node based on the accuracy of the shared information. Such a mechanism has therefore been proposed in this paper. The proposed system utilises the results of any false message detection method to generate and share feedback in the network, this feedback is then collected and filtered to remove potentially malicious feedback. A dynamic reputation score system is produced for each node. The reputation system has been experimentally validated and proved to have high accuracy in the detection of malicious nodes sending false information and is robust or negligibly affected in the presence of spurious feedback.