工业控制网络入侵检测的BP神经网络优化方法
|
摘要
针对工业控制系统入侵检测模型对各类攻击的检测率和检测效率不高的问题,提出一种Ada Boost算法优化BP神经网络的入侵检测模型.首先利用主成分分析法对原始数据集进行预处理,消除其相关性;其次利用Ada Boost算法对训练样本的权重进行不断调整,从而获得BP神经网络最优权重和阈值;最后再通过Ada Boost算法将BP弱分类器组合成BP强分类器,从而实现工业控制系统的异常检测.实验结果表明该方法在对各攻击类型的检测率和测试时间明显优于其他算法模型.
Aiming at the problem of low detection rate of typical attack types and low detection efficiency in industrial control system intrusion detection model, an intrusion detection model optimized by BP neural network based on AdaBoost algorithm is proposed. Firstly, the original data set is preprocessed by PCA to eliminate its correlation; Secondly, the AdaBoost algorithm is used to continuously adjust the weight of the training samples,thereby obtaining the optimal weight and threshold of BP neural network.Finally, the BP weak classifier is combined into a BP strong classifier by AdaBoost algorithm, and the anomaly detection of the industrial control system is realized. The experimental results show that the methods are superior to other algorithm models in detecting the attack types and test time.
Aiming at the problem of low detection rate of typical attack types and low detection efficiency in industrial control system intrusion detection model, an intrusion detection model optimized by BP neural network based on AdaBoost algorithm is proposed. Firstly, the original data set is preprocessed by PCA to eliminate its correlation; Secondly, the AdaBoost algorithm is used to continuously adjust the weight of the training samples,thereby obtaining the optimal weight and threshold of BP neural network.Finally, the BP weak classifier is combined into a BP strong classifier by AdaBoost algorithm, and the anomaly detection of the industrial control system is realized. The experimental results show that the methods are superior to other algorithm models in detecting the attack types and test time.
引文
[1]尚文利,安攀峰,万明,等.工业控制系统入侵检测技术的研究及发展综述[J].计算机应用研究,2017,34(2):328-330.SHANG Wenli,AN Panfeng,WAN Ming,et al.Summary of research and development of industrial control system intrusion detection technology[J].Application Research of Computers,2017,34(2):328-330.
[2]杨安,孙利民,王小山,等.工业控制系统入侵检测技术综述[J].计算机研究与发展,2016,53(9):2 039-2 054.YANG An,SUN Liming,WANG Xiaoshan,et al.Summary of intrusion detection technology for industrial control systems[J].Computer Research and Development,2016,53(9):2 039-2 054.
[3]陈佳威.基于神经网络的入侵检测[J].无线互联科技,2017.23(11):154-156.CHEN Jiawei.Neural network based intrusion detection[J].Wireless Internet Technology,2017,23(11):154-156.
[4]沈夏炯,王龙,韩道军,等.人工蜂群优化的BP神经网络在入侵检测中的应用[J].计算机工程,2016,42(2):190-194.SHEN Xiajiong,WANG Long,HAN Daojun,et al.Application of artificial bee colony optimized BP neural network in intrusion detection[J].computer engineering,2016,42(2):190-194.
[5]杨云峰,唐凤仙,周来恩,等.改进遗传算法优化神经网络的入侵检测研究[J].河池学院学报,2017,37(2):77-83.YANG Yunfeng,TANG Fengxian,ZHOU Laien,et al.Research on intrusion detection based on improved genetic algorithm optimized neural network[J].Journal of Hechi University,2017,37(2):77-83.
[6]梁辰,李成海.PCA-BP神经网络入侵检测方法[J].空军工程大学学报,2016,32(6):93-96.LIANG Cheng,LI Chenhai.PCA-BP neural network intrusion detection method[J].Journal of Air Force Engineering University,2016,32(6):93-96.
[7]陈万志,李东哲.结合白名单过滤和神经网络的工业控制网络入侵检测方法[J].计算机应用,2018,38(2):363-369.CHEN Wanzhi,LI Dongzhe.Industrial control network intrusion detection method combining white list filtering and neural network[J].Journal of Computer Application,2018,38(2):363-369.
[8]陈万志,唐雨,张静,等.工业控制网络异常通信检测的改进鱼群算法优化方法[J].计算机应用研究,2018,36(08):323-328.CHEN Wanzhi,TANG Yu,ZHANG Jing,et al.Improved fish school algorithm optimization method for abnormal communication detection in industrial control networks[J].Application Research of Computers,2018,36(8):323-328.
[9]JENSEN P T.Detection and characterization of port scan attacks[J].Key Engineering Materials,2014,602-603(3):93-96.
[10]牛雷,孙忠林.PCA-AKM算法及其在入侵检测中的应用[J].计算机科学,2018,45(2):226-229.NIU Lei,SUN Zhonglin.PCA-AKM algorithm and its application in intrusion detection[J].Computer Science,2018,45(2):226-229.
[11]崔珂,付丽霞,张勇,等.基于DE优化的BP神经网络在入侵检测中的应用[J],软件导刊,2018,17(7):173-176.CUI Ke,FU Lixia,etal.Application of BP neural network based on de optimization in intrusion detection[J].Software guide,2018,17(7):173-176.
[12]董超,周刚,刘玉娇,等.基于改进的Ada Boost算法在网络入侵检测中的应用[J].四川大学学报,2015,11(28):1 225-12 29.DONG Chao,ZHOU Gang,LIU Yujiao,et al.Application of improved adaboost algorithm in network intrusion detection[J].Journal of Sichuan University,2015,11(28):1 225-1 229.
[13]ZHANG Tao,LI Baodian.Loan prediction model based on AdaBoost and PSO-SVM[J].Computer Engineering(NCCE 2018),2018,26(5):748-754
[14]曹建芳,陈俊杰,李海芳,等.基于Ada Boost-BP神经网络的图像情感分类方法研究[J],山西大学学报,2017,10(3):331-337.CAO Jianfang,CHEN Junjie,LI Haifang,et al.Research on image sentiment classification based on adaboost-bp neural network[J].Journal of Shanxi University,2017,10(3):331-337.
[15]王军,费凯,程勇,等.基于改进的Ada Boost-BP模型在降水中的预测[J].计算机应用,2017,10(9):2 689-2 693.WANG Jun,FEI Kai,CHENG Yong,et al.Prediction of precipitation based on improved adaboost-bpmodel[J].Computer Application,2017,10(9):2 689-2 693.
[2]杨安,孙利民,王小山,等.工业控制系统入侵检测技术综述[J].计算机研究与发展,2016,53(9):2 039-2 054.YANG An,SUN Liming,WANG Xiaoshan,et al.Summary of intrusion detection technology for industrial control systems[J].Computer Research and Development,2016,53(9):2 039-2 054.
[3]陈佳威.基于神经网络的入侵检测[J].无线互联科技,2017.23(11):154-156.CHEN Jiawei.Neural network based intrusion detection[J].Wireless Internet Technology,2017,23(11):154-156.
[4]沈夏炯,王龙,韩道军,等.人工蜂群优化的BP神经网络在入侵检测中的应用[J].计算机工程,2016,42(2):190-194.SHEN Xiajiong,WANG Long,HAN Daojun,et al.Application of artificial bee colony optimized BP neural network in intrusion detection[J].computer engineering,2016,42(2):190-194.
[5]杨云峰,唐凤仙,周来恩,等.改进遗传算法优化神经网络的入侵检测研究[J].河池学院学报,2017,37(2):77-83.YANG Yunfeng,TANG Fengxian,ZHOU Laien,et al.Research on intrusion detection based on improved genetic algorithm optimized neural network[J].Journal of Hechi University,2017,37(2):77-83.
[6]梁辰,李成海.PCA-BP神经网络入侵检测方法[J].空军工程大学学报,2016,32(6):93-96.LIANG Cheng,LI Chenhai.PCA-BP neural network intrusion detection method[J].Journal of Air Force Engineering University,2016,32(6):93-96.
[7]陈万志,李东哲.结合白名单过滤和神经网络的工业控制网络入侵检测方法[J].计算机应用,2018,38(2):363-369.CHEN Wanzhi,LI Dongzhe.Industrial control network intrusion detection method combining white list filtering and neural network[J].Journal of Computer Application,2018,38(2):363-369.
[8]陈万志,唐雨,张静,等.工业控制网络异常通信检测的改进鱼群算法优化方法[J].计算机应用研究,2018,36(08):323-328.CHEN Wanzhi,TANG Yu,ZHANG Jing,et al.Improved fish school algorithm optimization method for abnormal communication detection in industrial control networks[J].Application Research of Computers,2018,36(8):323-328.
[9]JENSEN P T.Detection and characterization of port scan attacks[J].Key Engineering Materials,2014,602-603(3):93-96.
[10]牛雷,孙忠林.PCA-AKM算法及其在入侵检测中的应用[J].计算机科学,2018,45(2):226-229.NIU Lei,SUN Zhonglin.PCA-AKM algorithm and its application in intrusion detection[J].Computer Science,2018,45(2):226-229.
[11]崔珂,付丽霞,张勇,等.基于DE优化的BP神经网络在入侵检测中的应用[J],软件导刊,2018,17(7):173-176.CUI Ke,FU Lixia,etal.Application of BP neural network based on de optimization in intrusion detection[J].Software guide,2018,17(7):173-176.
[12]董超,周刚,刘玉娇,等.基于改进的Ada Boost算法在网络入侵检测中的应用[J].四川大学学报,2015,11(28):1 225-12 29.DONG Chao,ZHOU Gang,LIU Yujiao,et al.Application of improved adaboost algorithm in network intrusion detection[J].Journal of Sichuan University,2015,11(28):1 225-1 229.
[13]ZHANG Tao,LI Baodian.Loan prediction model based on AdaBoost and PSO-SVM[J].Computer Engineering(NCCE 2018),2018,26(5):748-754
[14]曹建芳,陈俊杰,李海芳,等.基于Ada Boost-BP神经网络的图像情感分类方法研究[J],山西大学学报,2017,10(3):331-337.CAO Jianfang,CHEN Junjie,LI Haifang,et al.Research on image sentiment classification based on adaboost-bp neural network[J].Journal of Shanxi University,2017,10(3):331-337.
[15]王军,费凯,程勇,等.基于改进的Ada Boost-BP模型在降水中的预测[J].计算机应用,2017,10(9):2 689-2 693.WANG Jun,FEI Kai,CHENG Yong,et al.Prediction of precipitation based on improved adaboost-bpmodel[J].Computer Application,2017,10(9):2 689-2 693.