基于量子门线路神经网络的信息安全风险评估  

Risk Assessment of Information Security Based on Quantum Gate Circuit Neural Networks

作　　者：周超[1] 潘平[1] 黄亮[1] ZHOU Chao;PAN Ping;HUANG Liang(College of Computer Science and Technology,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学计算机科学与技术学院,贵阳550025

出　　处：《计算机工程》2018年第12期39-45,共7页Computer Engineering

基　　金：贵州省教育厅自然科学研究项目"大数据聚类的量子算法研究"(黔教合KY(2015)367号)

摘　　要：信息安全风险评估是对不确定的和随机的潜在风险进行综合评价的过程,目的是有效抑制、转移系统风险。在分析信息系统安全要素与保障体系的基础上,构建基于信息资产的信息系统安全风险评估模型,通过风险评估指标体系,得到实际检测的评估对象属性。利用一组量子门线路构建神经网络模型,将评估对象属性样本归一化处理结果作为网络输入并用量子位表示,经量子旋转门进行相位旋转并控制量子位的翻转。经过网络处理后,得到评估对象的综合风险。实验结果表明,与传统BP神经网络相比,该方法能够实现信息系统的风险评估,具有更优的收敛性能与更精确的风险预测能力,可为风险管理提供可靠的理论依据。Information security risk assessment is a process of comprehensive evaluation of uncertain and random potential risks in order to effectively suppress and transfer systemic risks.On the basis of analyzing the security elements and security system of information system,the information security risk assessment model based on information assets is constructed.Through the risk assessment index system,the attributes of the evaluation objects actually are obtained.The neural network model is constructed by using a set of quantum gate circurit.The normalized processing result of the attribute samples of the evaluation object is used as the network input and expressed by the quantum bit.The phase rotation is performed by the quantum revolving gate and the quenching of the quantum bit is controlled.After being processed by the network,the comprehensive risk of the evaluated object is obtained.The experimental results show that compared with the traditional BP neural network,this method can realize the risk assessment of information systems,with better convergence performance and more accurate risk prediction ability,which can provide a reliable theoretical basis for risk management.

关 键 词：信息系统 量子门线路 信息资产 风险评估 神经网络 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]