The network security situation predicting technology based on the small-world echo state network
Fenglan Chen, Yongjun Shen, Guidong Zhang, Xin Liu · 2013
Network security model is a complex nonlinear system, and the network security situation value possesses the chaotic characters. The predictability of these situation values is of great significance for network security management. This paper proposes a novel prediction method, which is based on the echo state networks (ESNs) with small-world property. We can utilize this method to predict the network security situation after training and testing the acquired historical attack records. Verified by simulation results, the method has a higher prediction accuracy and speed compared with the conventional ESNs. Therefore it can reflect the network security situation in the future timely and accurately. We believe that this achievement will provide some practical guides for network administrators to supervise the network status.