An Ensemble Intrusion Detection Method for Train Ethernet Consist Network Based on CNN and RNN

The train Ethernet Consist Network (ECN) undertakes the task of transmitting critical control instructions. With increasing interactions between network and outside environment, masses intrusions are threatening data security railway vehicles. intrusion detection system has been proved to be an efficient method detect attacks. In this paper, a novel ensemble is proposed defense attacks against ECN, in particular IP Scan, Port Denial Service (DoS) Man Middle (MITM). Thirty-four features different protocol contents extracted from raw generated our ECN testbed form specific dataset. A imaging temporal sequence building designed optimize Six base classifiers built based on several typical convolutional neural networks recurrent networks: LeNet-5, AlexNet, VGGNet, SimpleRNN, LSTM GRU. dynamic weight matrix voting integrate all classifiers. evaluated experiment results show that outstanding ability aggregate advantages achieves superior performance with accuracy 0.975.