针对网络安全态势感知问题,该文对多种已有态势感知方法进行比较和分析,提出了一种基于神经网络的网络安全态势感知方法。首先,设计了一种基于BP(back propagation)神经网络的网络安全态势评估方法。然后,为了解决态势要素与评估结果之间的不确定性及模糊性问题,提出了一种基于RBF(radical basis function)神经网络的网络安全态势预测方法,利用RBF神经网络找出网络态势值的非线性映射关系,采用自适应遗传算法对网络参数进行优化并感知网络安全态势。通过真实网络环境的实验验证了该文提出方法在网络安全态势感知中的可行性和有效性。
A situational awareness method was developed based on neural networks to improve network security situational awareness. A network security situational awareness method is proposed using BP (back propagation) neural network. A network security situation forecasting method using an RBF (radical basis function) neural network is used to solve for the uncertainty in the relation between the situation factor and the e,Jaluation result. The analysis gives a non-linear mapping between the network situational values, with the network parameters optimized by an adaptive genetic algorithm. Tests with a real network environment show that this method effectively improves network security situation awareness.