用于网络安全态势预测的SAGPSO-SVM模型研究

Research on SAGPSO-SVM Model for Network Security Situation Prediction

网络安全态势预测精度不足,导致主动防御不及时。本文提出一种基于模拟退火与高斯扰动的粒子群算法(SAGPSO)与支持向量机(SVM)结合的预测模型,首先在传统粒子群(PSO)算法基础上引入模拟退火的思想,避免局部极值,对较优个体进行高斯扰动,然后利用该算法的全局收敛性强、收敛速度快和精确度高的特点对SVM参数进行优化,最后运用获取的模型参数进行预测,从而提高预测精度,并将此模型的预测结果与PSO-SVM和SAPSO-SVM预测模型的预测结果进行对比。结果表明,SAGPSO-SVM是一个预测精度高而且能够更加准确的描述网络安全态势变化趋势的预测模型。

The accuracy of the network security status is poorly predicted,resulting in unpredictable active defense.A prediction model based on simulated annealing and Gaussian disturbance particle swarm optimization(SAGPSO)com-bined with SVM is proposed in this paper.Firstly,the idea of simulated annealing is introduced to avoid local extre-mum based on the traditional PSO algorithm,then Gaussian disturbance is performed on the better individuals.Further-more,the SVM parameters are optimized by the algorithm with the advantage of global convergence,fast convergence and high accuracy.Finally,the obtained model parameters are used for prediction,thereby improving the accuracy of the prediction.And compare the prediction results of this model with the prediction results of PSO-SVM and SAP-SO-SVM prediction models.The results show that SAGPSO-SVM is a predictive model with high prediction accuracy and more accurate description of the trend of network security situation changes.