基于动态贝叶斯网络的复杂网络攻击方法研究

Research of complex network node attack basded on dynamic Bayesian network

在制定网络攻击策略时,目标网络信息存在不确定性,攻击方缺乏综合、可靠、实时的攻击依据,难以达到攻击效果,为此提出一种科学的复杂网络攻击方法。对网络攻击中攻击方收益、损耗、代价和遇到的风险进行分析,建立指标体系,利用动态贝叶斯网络对网络节点的攻击效果进行综合评估,克服了传统节点重要度评估方法依靠网络拓扑单一指标或者对目标节点进行静态评估的缺点。仿真实验表明该方法在攻击时综合了更多节点关系和观测信息,避免了依靠静态评估手段实施攻击时实际攻击效果与理论期望的差距,同时在攻击精度上更加准确,攻击效能更高。

The uncertain attack information and the lack of a reliable basis for command and decision-maker in the formulation of the attack plan in complex network attacks,lead to the effect of attacks difficult to achieve mission objectives when implementing action.For this situation,this paper presents a scientific complex network attack effect evaluation method.By analyzing income,cost of attack,loss of the attacker,attacker risks encountered network attacks,the paper establishes the corresponding index system,through DBN method for a comprehensive assessment of the effect of attacking nodes in the network,effectively overcomes the traditional node selection method relying on a single indicator of network topology,the results show that the fuzzy dynamic model synthesizes more nodes connection and informations of observed data and can overcome the static evaluation method to carry out attacks in the gap between the actual attack and theory of expectation effect while making more precision and higher performance on attack.