Vehicular network anomaly detection based on 2-step deep learning framework

Abstract

Intelligent Transportation System (ITS) is one of the newest technologies in the transportation sector that will give hope for better driving safety. Not only in terms of driving safety, but ITS will give also hope for driving comfort. Smart vehicles perchance better versatile to the street circumstances through trade data among vehicles. In case, they can maintain a strategic distance from activity blockage, perilous deterrents, or see activity mishaps prior. The innovation which is meticulously associated with the security of the driver must get extraordinary consideration. V2V-Vehicle-to-Vehicle connection can undermine impedance and indeed attack or anomaly. Many studies have been carried out to address this problem. The primary step is to reinforce the system's capacity to identify anomalies on Vehicular Network. Further, the growing development of machine learning seems to bring hope to support these steps. Within the proposed method, the original of our approach consists in utilizing 2-Step of anomaly detection. This framework is utilizing two classifiers machine learning from two altered preparing data-sets. We appear that the proposed method can make strides essentially attack detection achievement, compared to arrangements depending on a single detection step.