一种拒绝服务攻击的离散概率分布预测模型

首先对聚类问题进行了描述,然后采用遗传算法实现聚类方法的优化;在优化聚类的基础上得到样本数据中网络流量-攻击数量映射关系的多种类别,建立了拒绝服务攻击的多个预测子模型,并基于Bayesian方法推导出各子模型输出对应取值概率计算方法,得到拒绝服务攻击离散概率分布预测模型.

A Prediction Model of DoS Attack's Distribution Discrete Probability

The cluster problem is described first, and then a genetic algorithm is adopted to implement the optimization of the cluster method. Based on the optimized cluster method in the sample data, the various categories of the relation between the traffic and the attack amount are obtained and several prediction sub-models about DoS attack are built. According to the Bayesian method, the discrete probability calculation method about each sub-model is deduced and the prediction model of DoS attack's distribution d...