针对网络系统可生存性的有效评估问题,对指标体系的构建,指标赋权及综合评估方法进行了研究。提出了一种机制关联的网络可生存性指标体系构建方法,从可生存性的4个重要属性角度构建了层次化的指标体系。依据该指标体系提出了一种新的组合赋权方法,由层次分析法计算主观权重,粗糙熵方法计算客观权重,并给出了关联权重的概念和计算方法,构建离差最小化目标优化函数将3种权重联合得到组合权重,建立了基于隶属云的模糊综合评估模型,将隶属云替代原来的确定性隶属度函数。仿真结果表明,该评估方法可对网络系统可生存性进行准确有效的评估。
Aiming at the problem of evaluating network system survivability efficiently, the construction of index systems, weighting of indexes and synthetic evaluation method are studied. A constructing method of the network survivability index system by way of mechanism-associating is proposed, the hierarchy index system is constructed from the angle of survivability's four important attributes. A new combinatorial weighting method is given based on the index system. The subjective weight is calculated using analytic hierarchy process. The objective weights using rough entropy is calculated. The concept and calculating method of the associated weight are given. The three weights are combined into combinatorial weight using a minimum object optimization function. The fuzzy synthetic evaluation model based on cloud membership is established. The primary determinate membership function is replaced by cloud membership. The simulation results show that this evaluation method could evaluate the network system survivability accurately and efficiently.