MVR Delay: Cooperative Light Weight Detection and Prevention of False Emergency Message Dissemination in VANET

The fake emergency message propagation is one of the major problems in VANET networks where attackers propagate false emergency messages for their selfish needs and disrupt the reliability of VANETs. Emergency messages are intended to be delivered fast and the network should be optimized for faster delivery of emergency messages. A complex authentication procedure disturbs the utilities of emergency messaging in VANET. Towards this end, the emergency message propagation must pass through light weight verification without affecting the speed of propagation of genuine messages. Since the attack behavior changes over a period of time, the detection system must consider temporal correlation between attack characteristics over a period of time. This work proposes a secure path for emergency message dissemination in VANET networks wherein the authenticity of message is verified using Long short-term memory (LSTM) classifier using the temporal correlation between the events. The emergency message along the path is verified and fake messages are dropped by the Road side units (RSU). Simulation of proposed scheme shows more than 96% accuracy in filtering fake emergency messages and at the same time it does not introduces any significant delay in message delivery.