Network Data Analysis and Anomaly Detection Using CNN Technique for Industrial Control Systems Security

Abstract

Industrial control system (ICS) security is an important topic in context of Industry 4.0. Due to limited industrial network data and insufficient data analysis, anomaly detection for ICS is hardly to implement to enforce ICS security. The paper analyzes the traditional IT network data and ICS network data, bridges the common knowledge of network flow data feature of the both networks, and transfers the anomaly detection knowledge of traditional IT network into ICS network by means of a designed convolutional neural network (CNN) mechanism. Specific experiments validate the accuracy and reliability of the proposed CNN mechanism.