Vehicular ad hoc networks (VANETs) are most promising technology for smart transportation systems to provide road safety. However, VANETs are more vulnerable to malicious nodes and susceptible to a number of security attacks. In order to provide security against these attacks, most of the researchers have presented cryptography based schemes. These schemes ensure legitimate sender but cannot prevent broadcasting bogus message from legitimate sender. Hence, many researchers have proposed trust management based schemes to address internal attacks. However, these schemes encounter with complex iterations and computation overhead. To address these issues, in this paper, a deep learning based driver classification and trust computation (DL-DCTC) scheme is proposed. The sequential Deep Neural Network models are presented to calculate reward-points based on driver behaviour and classify fraudulent and non- fraudulent message / driver. The trust of a vehicle is computed using its reward-points earned during Vehicle to Vehicle (V2V) communications. Extensive simulation results are presented to validate effectiveness of proposed DL-DCTC scheme. The performance analysis showed that proposed scheme's performance is comparatively improved by reduction in computation overhead.
A Deep Learning Based Driver Classification and Trust Computation in VANETs
2019-09-01
526817 byte
Conference paper
Electronic Resource
English
Towards a trust management for VANETs
IEEE | 2014
|Decentralized Trust Management System for VANETs
Springer Verlag | 2022
|A trust propagation scheme in VANETs
IEEE | 2009
|A Trust Propagation Scheme in VANETs
British Library Conference Proceedings | 2009
|