Network intrusion detection method based on variational auto-encoder and depth echo state

Abstract

The invention discloses an industrial intrusion detection method based on a variational auto-encoder and a deep echo state network, and relates to an industrial control network intrusion detection method. The method comprises the following steps: standardizing a captured data set, dividing a training set and a test set, considering a sample imbalance problem, generating a minority class attack sample by utilizing a probability generation characteristic of a variational auto-encoder, sending the processed data into a depth echo state network model to be trained, and obtaining a deep echo state network model to be trained. The echo state network abandons a back propagation mechanism, so that the training time is greatly shortened, and potential features in the data can be effectively extracted through the stacking reserve pool. Results show that the method not only achieves relatively short training time consumption, but also improves the detection performance such as accuracy, precision and detection rate to a certain extent, is more suitable for industrial environments, and in addition, can still achieve a good detection effect when a sample imbalance problem exists in an intrusion detection data set.

Description

technical field [0001] The invention relates to an industrial control network intrusion detection method, in particular to an industrial intrusion detection method based on a variational autoencoder and a deep echo state network. Background technique [0002] Hackers' cyberattacks on industrial control systems (ICS) threaten national security. Industrial control networks (ICN) are an important part of industrial control systems. The development of information technology brings convenience to people, but also security. Challenges. Early industrial control systems only focused on practicability and stability, and must ensure long-term uninterrupted operation. Interrupting these systems just for safety updates will cause economic losses and affect work efficiency. For example, in urban power systems, a short power outage can It may affect the daily life of tens of thousands of people. The industrial control system is the core of the national infrastructure. Once it is attacked...

AI Technical Summary

Problems solved by technology

[0005] The object of the present invention is to provide a kind of industrial intrusion detection method based on variational autoencoder and deep echo state network, this method utilizes variational autoencoder (Variational Auto-Encoders, VAE) to collect the minority class in the data set VAE is a deep generative model, and its ability to reconstruct the probability density distribution of the original input samples can also be used to generate abnormal traffic; to solve the problem of low detection rate due to the complex structure of the model, this method is based on the deep echo state network. Network (Deep Echo State Network, DESN) intrusion detection method (D-IDS), ESN greatly shortens the training time due to the abandonment of the back propagation mechanism, and can effectively extract the potential features in the data by stacking the reserve pool

Images