基于卡尔曼滤波的网络安全预测

Prediction for Network Security Situation Based on Kalman Filter

针对现有网络安全态势预测模型往往只能对离线数据进行预测,无法根据历史信息来对当前数据进行在线实时预测的问题,提出了一种基于卡尔曼滤波的网络安全态势预测方法.首先,介绍了卡尔曼滤波的基本原理和相应的状态方程和预测方程,然后设计了基于卡尔曼滤波的网络安全态势在线预测算法,算法采用最小二乘方法来求解预测方程中的权值矩阵和噪声参数,从而获得了相应的预测表达式,可以实现对网络安全态势的实时预测.为了验证所提模型的预测效果,将其应用于离线数据集和在线实时数据中,并与其它方法进行了比较.仿真结果表明:所提模型能准确有效地对网络安全态势进行预测,具有预测精度高的优点,和其他方法相比,具有较大的优越性.

Aiming at the existing prediction model for network security situation only can concerns the off-line data set,it cannot able to predict according to the on-line data in time,yet a prediction method based on Kalman filter is proposed.Firstly,the basic principles,the state functions and the prediction functions are introduced,then the prediction algorithm based on Kalman filter is proposed,where the least squares method is used to learn the weight matrix and the noise parameter,so the corresponding expression can be obtained and the network security situation can be predicted in time.To verify the prediction effect of the proposed model,it is applied to the off-line data set and the on-line data set,the simulation result shows it can predict the network security situation accurately.Compared with the other methods,it has a relative large priority.