Route Intrusion Detection Based on Long Short Term Memory Recurrent Neural Network

With the rapid development and extensive use of the Internet, network security is the key point of the current research of computer researchers. The router is the key interconnection equipment in the Internet system, and it is necessary to improve the security of the router itself. The routing protocol is used in the packet switching network. Computer communication is implemented by each switching node in the network to forward data packets. At present, most of the router intrusion detection systems are based on the machine learning method. This paper proposes a Long Short Term Memory (LSTM) architecture applied to the Recurrent Neural Network (RNN), and uses the KDD Cup 1999 data set to train the IDS model. Through performance testing, the ability to efficiently and accurately classify network traffic into attacks and normal capabilities.