Intrusion-detection model integrating anomaly with misuse for space information network

Abstract

Abstract:In recent years, following the development of space commutation, space information has become a critical part in space information network and will play a very significant role in winning future information war. A space information network with characteristics such as complex structure, special communication requirement, long delay, dependence on remote maintenance, and fragile ecological environment contains enormous security risks. Therefore, ensuring space information network safety is important. Intrusion-detection model as an important part of a network security system becomes a hot issue in space network security. We propose an intrusion-detection method that integrates anomaly with misuse, which supports automatic updates from a remote ground, and design a distributed intrusion-detection model of space information network.

Online Network traffic anomaly detection method combining OS-ELM and SADE

“…(3) Reduce the dimension of feature vectors in the sample data set using SDAE network. (4) The processed data set is used to train the OS-ELM classifier which is used to detect the anomaly network. The OS-ELM training process can be subdivided into the following steps.…”

“…Any unknown or illegal user can access the LAN through the network access point, but the existing firewall strategy cannot meet the requirements of high security. Therefore, anomaly detection method becomes a reasonable supplement to firewall, and only real-time IDS can maximize the detection of network abnormal behaviors [2][3][4][5]. In recent years, scholars have proposed a variety of anomaly detection methods based on data mining.…”