A New Factor State Space Model for SCADA Network Attack and Defense
A New Factor State Space Model for SCADA Network Attack and Defense
复制标题
DOI：
10.14257/ijsia.2014.8.6.27
复制
发表时间：
2014-11
影响因子：
--
通讯作者：
Li Yang;Ling Wang;Xinyu Geng;Xiedong Cao
中科院分区：
文献类型：
--
作者：
Li Yang;Ling Wang;Xinyu Geng;Xiedong Cao
针对监控与数据采集（SCADA）系统中的安全问题，提出了一种基于因素状态空间的SCADA网络攻防因素网络模型。结合因子空间理论，给出了基于因子状态空间的因子神经元的形式化描述。在分析和表达网络攻防因素的基础上，提出了基于变权的因素神经元模型，提出了一种基于模糊神经网络的SCADA网络安全防御体系结构模型。通过引入因子空间藤，利用攻击仿真实验验证了该方法在网络攻防知识推理中的可行性。实验结果表明，该方法能有效提高对不同攻击的识别率。基于因子状态空间的因子神经网络能有效解决网络攻防系统中知识推理和表达的复杂性，为解决类似应用提供了一种新的方法。
To solve the security problem in the supervisor control and data acquisition (SCADA), a new factor network model of SCADA network attack and defense based on factor state space is presented. Combining with factor space theory, formal descriptions of factor neurons based on factor state space are developed. On the basis of analysis and expression of network attack and defense factors, factor neuron model based on variable weight is proposed and a FNN-based security defense architecture model for SCADA network is put forward. For illustration, by introducing factor space canes, an attack simulation experiment is utilized to show the feasibility of the proposed method in solving network attack and defense knowledge reasoning. Experimental results indicate that the proposed method can effectively improve recognition rate of different attacks. Factor neuron network based on factor state space can effectively solve complexity of knowledge reasoning and expression in network attack and defense system and provides a new method for solving similar application.