摘要：Air Traffic Management (ATM)system is enticing targets of cyber-attacks since 9.11 event, and the securitysituation of ATM information system is closely related to the safety flight ofair transportation. In this paper, an approach of security evaluation for ATMsystem is proposed based on artificial neural network (ANN). The proposed approachcombines neural networks with Bayesian regularization to simulate ATM system inANN. An indicator system of security situation for ATM system is established asthe input of ANN, and the output is system security level, which is obtained bytraining the ANN with the daily recoded data of ATM system. Experimentalresults show that the proposed method has a certain precision and practicality.
