A Prediction Approach To Cyber Attack Paths Of Urban Rail Transit Control Systems

With the wide application of communication-based train control(CBTC)technology,the level of informationization and automation of urban rail transit has been greatly improved.However,the utilization of computers,communication,and control technologies has made the cyber-security of CBTC systems increasingly prominent.Due to safety critical characteristics of the CBTC system,the existing safety assurance technology can restrict the dissemination of cyber-security risks to a certain extent.It is necessary to implement the dynamic and quantitative evaluation and prediction on the evolution process of cyber security risks,which is important to perform the efficient defense and enhance the security level of the CBTC system.Based on architecture characteristics and operation mechanisms of CBTC systems,the attack path library is built through the penetration attack graph,according to vulnerability analysis.With the attack-defense game,an attack path prediction method for CBTC system is developed,which can provide the technical basis for the research on the evolution mechanism of cyber-security risks.The main work of the thesis is as follows:A method for the attack path library of the CBTC system based on the penetration attack graph.According to the vulnerability distribution of the CBTC system and the characteristics of the redundant architecture,combining with the information flow direction of the system,the component vulnerability queue is established.With the hot standby node and the Boolean AND node,a penetration attack graph model of the CBTC system is built,which is efficient to obtain the attack path library with the fail-safe mechanisms of the CBTC systems.The attack path prediction method based on the attack-defense game.According to the multiple indicators of vulnerability,attack types and defense strategies which form a triple,a quantitative analysis method for attacks and defenses is proposed.On the basis,through the minmax criterion,an attack path prediction model is developed.The CBTC attack path prediction visualization system based on MVC architecture.According to the attack path prediction method proposed in this paper,the attack path search module,the attack path visualization module and the Nash equilibrium solution module are developed.The system can quickly find and demonstrate the most possible attacking object,which can provide assistant decisions.The attack-defense game model established in this paper can predict attack path very well,and the attack path prediction system developed in this paper can clearly and conveniently guide defense personnel to carry out key defenses.