Deep Learning Framework for Accurate Network Intrusion Detection in ITSs

According to the increase in the number of Internet-connected devices in Intelligent Transport Systems (ITSs), the network traffic carried by ITSs has significantly increased. Consequently, the number of new security threats and attacks has increased beyond the ability of exiting systems to solve. New technologies have been recently used to address the emerging security challenges in ITSs such as the use of machine learning techniques to predict new attacks and threats. In this paper, we propose an effective solution for this problem using deep learning to detect and classify the attacks in the widely used UNSW-NB15 dataset. A discriminating feature of our work is the use of a reduced set of features (only 20 out of 49) in the proposed model. Our experimental results show that the proposed model achieves remarkable accuracy despite the use of a subset of carefully selected features in the UNSW-NB15 dataset.