A New Intrusion Detection Feature Extraction Method Based on Complex Network Theory

Whether the most important features can be extracted to reduce the dimension of the features or not is crucial to improving the efficiency and performance of the Intrusion Detection System (IDS). In this paper, an intrusion detection feature extraction method based on the complex network theory and the MST algorithm is proposed. The method takes the features of the network connections as nodes of a scale-free model, then detects the clusters of the network and extracts the key nodes of the model. The extracted nodes can be used in the IDS to detect the existence of intrusions. The result shows that the detection rate of the method is almost 1 percent lower than that of the Principal Component Analysis (PCA) algorithm, but the efficiency is improved by 13 percent. At last, how to apply the method to the intrusion detection pattern match is discussed.