A secure and trusted context prediction for next generation autonomous vehicles

Abstract

To ensure better facilitation of vehicular services and improve driving safety in the Internet of Vehicles (IoV), context prediction among vehicles plays a very crucial role. However, as more malicious IoV devices get involved in the network, the context prediction accuracy shared among various servers may degrade severely. Existing schemes have used cryptographic mechanisms to securely and accurately identify malicious devices. However, time and the subsequent delay in identifying and rating the legitimate communicating IoV devices emerge as a crucial issue. Hence, to solve this critical problem, we put forth an efficient and reliable trust framework where trust and context prediction is achieved by Tidal Trust Mechanism (TTM) and Contract Theory (CT). TTM can successfully rate the degree of trust between the devices with a high level of accuracy, whereas CT can verify the context prediction reliably. The proposed mechanism based on TTM and CT ensures that trusted IoV devices are identified with high accuracy and verified reliably. The proposed framework is simulated over real-world data set in MATLAB for various performance metrics, such as altered records, accuracy prediction, response time, and utilities of IoV devices. Simulation results show that the proposed framework provides a significant improvement of approximately 87% in comparison to existing (baseline) approaches while analyzing the accuracy, record alteration, and resource utility among the devices in the network.