Real-Time Network Intrusion Detection System Based on Deep Learning

Abstract

Computer network is vulnerable to hackers, computer viruses, and other malicious attacks. As an active defense technology, intrusion detection plays an important role in the field of network security. Traditional intrusion detection technologies face problems such as low accuracy, low detection efficiency, high false positive rate, and inability to cope with new types of intrusions. To solve these problems, we propose a real-time network intrusion detection system based on deep learning, which uses big data technology, natural language processing technology and deep learning technology. Our main contributions are as follows: (1) Use Flume as the agent for log collection to realize real-time massive log collection, using Flink as real-time computation engine. (2) Aiming at the high dimensional problem of traffic data, a self-encoder-based intrusion detection dimension reduction method is proposed, and the intrusion detection data is preprocessed, including data cleaning, coding, extraction and integration, and normalization. (3) Propos a deep learning-based intrusion detection model, AE-AlexNet, which uses Auto-Encoder AlexNet neural network. The experimental results of the intrusion detection data set KDD 99 show that the accuracy of the AE-AlexNet, model is as high as 94.32%.