APP下载

基于MSCPO—SVR的网络态势预测模型

2018-12-05胡俊荷李芯蕊马双斌

网络空间安全 2018年4期

胡俊荷 李芯蕊 马双斌

摘 要:网络态势值是衡量网络安全的重要指标,其值的高低直接影响网络安全的优劣。能够有效地建立起网络态势预测模型,其意义对于预防网络安全事件的发生以及网络安全防护来讲非同寻常。通过对网络态势的理解和分析,可知影响网络态势的因素很多且关系比较复杂,难以建立较为准确的数学表达式来描述网络态势。因此,文章采用改进的粒子群算法优化支持向量回归机模型的参数,构建起网络态势与影响因素两者中的非线性映射关系,提出一种多种群的混沌粒子优化算法,建立MSCPO-SVR网络态势预测模型,并得出预测结果。最后,通过与BP神经网络和SVR预测模型进行对比,验证MSCPO-SVR预测模型的预测值与真实值的拟合性比较强。

关键词:计划行为理论;涉密人员;保密行为;保密行为意向

中图分类号:TP 309;TP 393 文献标识码:A

Abstract: The Network trend value is an important index of network security, and it affects the merits of Network security directly. Its significant to effectively establish the Network Security Situation Prediction Model (NSSP) for the precaution of network intrusion and protection of the network. Through the comprehension and analysis of network security situation, we know there are many factors that can affect the security situation of network and the relationship between them are very complex. Therefore its difficult for us to describe the net security situation in a specific mathematical formula. Under these circumstances, in this essay, an improved particle swarm optimization is used to optimize the parameters of the Support Vector Regression Model, and constructs a non-linear mapping relation between the network situation and the influencing factors. We also proposes a multi-population chaotic particle optimization algorithm, establishes the MSCPO-SVR network trend prediction model and obtains the predicted result. Finally, by comparing the MSCPO-SVR prediction model with BP neural network and SVR prediction model, we prove that the prediction value of MSCPO-SVR prediction model is more fitted with the real value.

Key words: situation prediction; particle swarm optimization algorithm; support vector regression; prediction model

1 引言

“互聯网+”的提出,使很多行业与互联网进行融合,网络成为了人们日常生活中密不可分的一部分。随着其规模日益壮大,网络已经普遍应用于各个领域。现在攻击者不仅对各种网络数据的窃取,而且也开始对一些关键性基础设施、政府、金融机构和能源行业等进行攻击,给社会和国家带来了一定的影响。

张伟丽[1]分析了我国等级保护存在的问题,并提出了进一步做好信息安全等级保护工作的建议;李连等[2]分析了云计算的基本特征及面临的安全问题,总结了目前国际上关于云计算安全问题的研究现状,讨论了云计算安全技术框架和关键技术;……

登录APP查看全文