面向云计算的网络安全风险预测模型的研究  

Research of network security risk prediction model based on cloud computing

作　　者：郭祖华[1] 李扬波[1] 徐立新[1] 张晓[2] 

机构地区：[1]河南机电高等专科学校计算机科学与技术系,河南新乡453003 [2]西北工业大学计算机学院,西安710072

出　　处：《计算机应用研究》2015年第11期3421-3425,共5页Application Research of Computers

基　　金：国家"863"计划资助项目(2013AA01A215);河南省教育厅科学技术重点研究项目(13A520221;14A520045);河南省教育科学"十二五"规划课题(2012-JKGHAC-0116*)

摘　　要：在云服务环境下,虚拟化对网络安全风险评估提出了更高要求。通过在云计算虚拟化平台上构造一个以虚拟组件为中心的系统风险评测模型,实现了对Web service的DDo S攻击的网络安全风险预测。该模型采用层次化分析法,实现了Web service的风险预测,提高了风险值的预测精度;采用RBF神经网络进行并行化预测,达到了多个节点实现快速实时预测的目的。实验测试表明,上述方法具有较好的风险评测效果,可为云计算系统的网络安全防御策略的制定提供科学的参考依据。In the cloud service environment, virtualization puts forward higher requirements to network security risk assessment. Through constructing a virtual component-centrie system risk evaluation model on the cloud computing virtualization platform, the paper implemented network security risk prediction for the Web service DDoS attack. This model used a hierarchical analysis method to achieve a risk prediction of Web service that improved the prediction accuracy of risk values. Meanwhile, the paper used of RBF neural network for prediction parallelization, which achieved the purpose of rapid real-time prediction with multiple nodes. Experimental tests show that the method has a better risk evaluation result, which can provide scientific basis for developing the network security defense strategy of cloud computing systems.

关 键 词：云计算 网络安全 风险预测 并行RBF 层次化分析法 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]