模糊动态贝叶斯网络防御态势感知模型

发布时间：2018-02-26 08:26

本文关键词： 动态贝叶斯网络网络空间防御态势感知态势评估　出处：《探测与控制学报》2017年03期 　论文类型：期刊论文

【摘要】：针对态势感知模型无法准确捕捉网络系统信息变化的弊端,提出了基于模糊动态贝叶斯网络的态势评估模型。通过对态势要素的状态进行模糊和概率化处理,构建了态势感知和态势评估两个子模型。模型归一化各级态势要素的状态并权值相加,按值划分态势区间,推演态势等级,形成态势处置方案为网络空间防御提供了辅助决策。将概率及观测数据输入模型中进行仿真评估,并与静态贝叶斯网络模型和基于Hopfield神经网络的评估模型进行结果对比,实验结果表明,基于模糊动态贝叶斯网络模型的评估结果综合了更多节点关系和观测信息,在对抗中能逐渐适应对抗强度,防御水平逐渐提高,随着对抗时间的延续,能较好地反映网络防御作战态势的变化规律,同时能有效对目标网络威胁等级进行分类,对态势结果和趋向进行准确分析预估。
[Abstract]:The situation assessment model based on fuzzy dynamic Bayesian network is proposed to solve the problem that the situation perception model can not accurately capture the information change of network system. The fuzzy and probabilistic processing of the state of situation elements is carried out. Two sub-models of situational awareness and situation assessment are constructed. The model normalizes the state and weights of each level of situation elements, divides the situation interval according to the value, and deduces the situation level. The formation of situation disposal scheme provides an auxiliary decision for cyberspace defense. The probability and observation data input model is simulated and evaluated, and the results are compared with the static Bayesian network model and the evaluation model based on Hopfield neural network. The experimental results show that the evaluation results based on fuzzy dynamic Bayesian network model synthesizes more node relations and observation information, and can adapt to the intensity of confrontation gradually, and the defense level increases gradually with the continuation of confrontation time. It can reflect the changing law of the network defense operational situation, and can effectively classify the target network threat level, and accurately analyze and predict the situation result and trend.
【作者单位】：江西科技学院信息工程学院;
【基金】：国家自然科学基金项目资助(61461029)
【分类号】：TP18;TP393.08

【相似文献】