An Improved Security Algorithm for VANET using Machine Learning

Abstract

Vehicular Ad-Hoc Networks (VANETs) have become a fascinating research area over the last decade due to the increasing number of Vehicles on road. A secure Intelligent Transportation System (ITS) ensures the safety of the passengers and the driver nevertheless the dynamic characteristics of it make it a challenging area in terms of real time implementation. This paper proposes an improved security algorithm for VANET, which is able to deal with the threats like Denial of Service Attack (DoS), Sybil and Replay. The proposed work uses Enhanced K-Mean method to create the clusters for various attacks and a hybrid approach using Support Vector Machine (SVM) and Feed-forward back propagation is used to test the classifier for its accuracy. The results show a significant improvement in terms of Throughput, Jitter and PDR. Finally, we highlight future direction and some open issues for further exploration.