Intrusion Detection System Using Machine Learning by RNN Method

Abstract

As computer networks continue to grow, network intrusions become more frequent, advanced, and volatile, making it challenging to detect them. This has led to an increase in illegal intrusions that current security tools cannot handle. NIDS is currently available and most reliable ways to monitor network traffic, identify unauthorized usage, and detect malicious attacks. NIDS can provide better visibility of network activity and detect any evidence of attacks and malicious traffic. Recent research has shown that machine learning-based NIDS, particularly with deep learning, is more effective in detecting variants of network attacks compared to traditional rule-based solutions. This proposed model that introduces novel deep learning methodologies for network intrusion detection. The model consists of three approaches: LSTM-RNN, various classifying methodology, and a hybrid Sparse autoencoder with DNN. The LSTM-RNN evaluated NSL-KDD dataset and classified as multi-attack classification. The model outperformed with Adamax optimizer in terms of accuracy, detection rate, and low false alarm rate.