融合杜鹃搜索的灰狼优化算法在网络入侵检测特征选择中的应用
|
摘要
针对当前网络入侵检测技术由于特征冗余引起的检测效率低和准确率低等问题,将一种融合杜鹃搜索的灰狼优化算法应用于网络入侵检测的特征选择中,旨在减少特征冗余,进而提高网络入侵检测的性能。首先,在每次迭代过程中采用杜鹃搜索算法中的莱维飞行机制对适应度值最好的3只灰狼的位置进行扰动,避免在搜索最优解的过程中陷入局部最优。然后,采用灰狼优化算法的更新机制来更新灰狼的位置信息,使狼群朝着猎物的方向聚集。最后,根据预先设定的概率值对狼群的位置进行随机更新,迫使狼群在不断逼近猎物的过程中,能随机地跳出局部最优,从而提高灰狼优化算法在网络入侵检测特征选择中的全局搜索能力。使用网络入侵检测NSL-KDD测试集进行验证实验,并与灰狼优化算法、杜鹃优化算法以及传统的信息增益算法从特征选择的角度进行对比,结果表明,将融合杜鹃搜索的灰狼优化算法应用于网络入侵检测的特征选择时,分类准确率及特征子集的选择都取得较好的效果。融合杜鹃搜索的灰狼优化算法在全局搜索能力方面有较为显著的提升,将其应用于特征选择中,可以有效地提高网络入侵检测的性能。
In order to solve the problem of low detection efficiency and low accuracy caused by feature redundancy in current network intrusion detection technology, an improved Grey Wolf Optimizer algorithm integrated with Cuckoo Search was applied to feature selection for network intrusion detection, which aimed at reducing the feature redundancy and improving the performance of network intrusion detection. First, the Levy flight mechanism in the Cuckoo Search algorithm was adopted to disturb the position of three grey wolves with the best fitness values, which avoided the local optimum in the process of searching the optimal solution. Then, the updating mechanism of the Grey Wolf Optimizer algorithm was utilized to update the location information of grey wolves, ensuing these wolves could gather in the direction of their prey. Finally, the location of the grey wolf group was randomly updated according to the predetermined probability value. Hence, the grey wolf group could randomly jump out of the local optimum in the process of continuous approximation of the prey, which improved the global-search ability of the Grey Wolf Optimizer algorithm in feature selection for network intrusion detection. The NSL-KDD dataset in network intrusion detection was used for the verification, and the Grey Wolf Optimizer algorithm, the Cuckoo Search algorithm and the traditional Information Gain algorithm were compared in the experiments from the viewpoint of feature selection. The experimental results showed that, when applying the improved Grey Wolf Optimizer algorithm integrated with Cuckoo Search to feature selection for network intrusion detection, it could achieve good results in both the classification accuracy and the selection of feature subsets. In summary, the improved Grey Wolf Optimizer algorithm integrated with Cuckoo Search had a relatively significant improvement in the ability of global search, and its application in feature selection could effectively improve the performance of network intrusion detection.
In order to solve the problem of low detection efficiency and low accuracy caused by feature redundancy in current network intrusion detection technology, an improved Grey Wolf Optimizer algorithm integrated with Cuckoo Search was applied to feature selection for network intrusion detection, which aimed at reducing the feature redundancy and improving the performance of network intrusion detection. First, the Levy flight mechanism in the Cuckoo Search algorithm was adopted to disturb the position of three grey wolves with the best fitness values, which avoided the local optimum in the process of searching the optimal solution. Then, the updating mechanism of the Grey Wolf Optimizer algorithm was utilized to update the location information of grey wolves, ensuing these wolves could gather in the direction of their prey. Finally, the location of the grey wolf group was randomly updated according to the predetermined probability value. Hence, the grey wolf group could randomly jump out of the local optimum in the process of continuous approximation of the prey, which improved the global-search ability of the Grey Wolf Optimizer algorithm in feature selection for network intrusion detection. The NSL-KDD dataset in network intrusion detection was used for the verification, and the Grey Wolf Optimizer algorithm, the Cuckoo Search algorithm and the traditional Information Gain algorithm were compared in the experiments from the viewpoint of feature selection. The experimental results showed that, when applying the improved Grey Wolf Optimizer algorithm integrated with Cuckoo Search to feature selection for network intrusion detection, it could achieve good results in both the classification accuracy and the selection of feature subsets. In summary, the improved Grey Wolf Optimizer algorithm integrated with Cuckoo Search had a relatively significant improvement in the ability of global search, and its application in feature selection could effectively improve the performance of network intrusion detection.
引文
[1]Mi Hong,Yang Xibei.Network anomaly detection model based on intrusion feature selection[J].Modern Electronics Technique,2017,40(12):69-71.[米洪,杨习贝.基于人侵特征选择的网络异常数据检测模型[J].现代电子技术,2017,40(12):69-71.]
[2]Ramakrishnan S,Devaraju S.Attack's feature selectionbased network intrusion detection system using fuzzy control language[J].International Journal of Fuzzy Systems,2017,19(2):316-328.
[3]Khammassi C,Krichen S.A GA-LR wrapper approach for feature selection in network intrusion detection[J].Computers&Security,2017,70:255-277.
[4]Liu Bailu,Yang Yahui,Shen Qingni.Method of intrusion early feature selection based on genetic algorithm[J].Journal of Chinese Computer Systems,2015,36(1):111-115.[刘白璐,杨雅辉,沈晴霓.一种基于遗传算法的人侵早期特征选择方法[J].小型微型计算机系统,2015,36(1):111-115.]
[5]Li An,Jiang Jiahe.A feature selection method for intrusion detection using the adaptive genetic algorithm[J].Applied Science and Technology,2016,43(3):49-53.[李安,江加和.基于自适应遗传算法的入侵检测特征选择方法[J].应用科技,2016,43(3):49-53.]
[6]Yuan Kaiyin,Fei Lan.Detection of network intrusion based on hybrid particle swarm optimization algorithm selection features[J].Journal of Jilin University(Science Edition),2016,54(2):309-314.[袁开银,费岚.混合粒子群优化算法选择特征的网络人侵检测[J].吉林大学学报(理学版),2016,54(2):309-314.]
[7]Mirjalili S,Mirjalili S M,Lewis A.Grey wolf optimizer[J].Advances in Engineering Software,2014,69:46-61.
[8]Lv Xinqiao,Liao Tianlong.Permutation flow-shop scheduling based on the grey wolf optimizer[J].Journal of Wuhan University of Technology,2015,37(5):111-116.[吕新桥,廖天龙.基于灰狼优化算法的置换流水线车间调度[J].武汉理工大学学报,2015,37(5):111-116.]
[9]Mao Senmao,Qu Kaiping,Chen Yixuan,et al.Grey wolf multi-objective optimizer for optimal carbon-energy combined-flow[J].The Journal of New Industrialization,2016,6(9):11-17.[毛森茂,瞿凯平,陈艺璇,等.基于灰狼多目标算法的电网碳-能复合流优化调度[J].新型工业化,2016,6(9):11-17.]
[10]Wang Tai,Xu Bin,Li Linguo,et al.A multi-threshold image segmentation algorithm based on discrete grey wolf optimization[J].Computer Technology&Development,2016,26(7):30-35.[王钛,许斌,李林国,等.基于离散灰狼算法的多级阈值图像分割[J].计算机技术与发展,2016,26(7):30-35.]
[11]Seth J K,Chandra S.Intrusion detection based on key feature selection using binary GWO[C]//Proceeding of International Conference on Computing for Sustainable Global Development.Piscataway,NJ:IEEE Press,2016:3735-3740.
[12]Emary E,Zawbaa H M,Grosan C,et al.Feature subset selection approach by gray-wolf optimization[C]//Proceedings of Afro-European Conference for Industrial Advancement.S witzerland:Springer,Cham:2014:1-13.
[13]陈昌帅,二进制灰狼优化算法的研究与分析[J].信息系统工程,2016,28(7):136-138.
[14]Yang X S,Deb S.Cuckoo search via Levy flights[C]//Proceedings of Nature&Biologically Inspired Computing,2009.Piscataway,NJ:IEEE Press,2009:210-214.
[15]Jiang Y,Liu X,Yan G,et al.Modified binary cuckoo search for feature selection:a hybrid filter-wrapper approach[C]//Proceedings of International Conference on ComputationalIntelligence and Security.Washington DC:IEEE Computer Society,2017:488-191.
[16]Aziz M A E,Hassanien A E.Modified cuckoo search algorithm with rough sets for feature selection[J].Neural Computing&Applications,2018,29(4):925-934.
[17]Raj A M J,Francis F S,Benadit P J,et al.Particle swarm and cuckoo search(PSCS)optimization based feature selection method to improve the web page classification[J].Journal of Advanced Research in Dynamical&Control Systems,2017,9(6):1125-1140.
[18]Xu H,Liu X,Su J.An improved grey wolf optimizer algorithm integrated with cuckoo search[C]//Proceedings of IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems:Technology and Applications.Piscataway,NJ:IEEE,2017:490-493.
[19]Rodrigues D,Pereira L A M,Almeida T N S,et al.BCS:A binary cuckoo search algorithm for feature selection[C]//Proceedings of IEEE International Symposium on Circuits and Systems.Piscataway,NJ:IEEE,2013:465-468.
[20]Canadian Institute for Cyber security.NSL-KDD dataset[EB/OL].(2009-07-08)[2018-07-05].http://www.unb.ca/cic/datasets/nsl.html.
[2]Ramakrishnan S,Devaraju S.Attack's feature selectionbased network intrusion detection system using fuzzy control language[J].International Journal of Fuzzy Systems,2017,19(2):316-328.
[3]Khammassi C,Krichen S.A GA-LR wrapper approach for feature selection in network intrusion detection[J].Computers&Security,2017,70:255-277.
[4]Liu Bailu,Yang Yahui,Shen Qingni.Method of intrusion early feature selection based on genetic algorithm[J].Journal of Chinese Computer Systems,2015,36(1):111-115.[刘白璐,杨雅辉,沈晴霓.一种基于遗传算法的人侵早期特征选择方法[J].小型微型计算机系统,2015,36(1):111-115.]
[5]Li An,Jiang Jiahe.A feature selection method for intrusion detection using the adaptive genetic algorithm[J].Applied Science and Technology,2016,43(3):49-53.[李安,江加和.基于自适应遗传算法的入侵检测特征选择方法[J].应用科技,2016,43(3):49-53.]
[6]Yuan Kaiyin,Fei Lan.Detection of network intrusion based on hybrid particle swarm optimization algorithm selection features[J].Journal of Jilin University(Science Edition),2016,54(2):309-314.[袁开银,费岚.混合粒子群优化算法选择特征的网络人侵检测[J].吉林大学学报(理学版),2016,54(2):309-314.]
[7]Mirjalili S,Mirjalili S M,Lewis A.Grey wolf optimizer[J].Advances in Engineering Software,2014,69:46-61.
[8]Lv Xinqiao,Liao Tianlong.Permutation flow-shop scheduling based on the grey wolf optimizer[J].Journal of Wuhan University of Technology,2015,37(5):111-116.[吕新桥,廖天龙.基于灰狼优化算法的置换流水线车间调度[J].武汉理工大学学报,2015,37(5):111-116.]
[9]Mao Senmao,Qu Kaiping,Chen Yixuan,et al.Grey wolf multi-objective optimizer for optimal carbon-energy combined-flow[J].The Journal of New Industrialization,2016,6(9):11-17.[毛森茂,瞿凯平,陈艺璇,等.基于灰狼多目标算法的电网碳-能复合流优化调度[J].新型工业化,2016,6(9):11-17.]
[10]Wang Tai,Xu Bin,Li Linguo,et al.A multi-threshold image segmentation algorithm based on discrete grey wolf optimization[J].Computer Technology&Development,2016,26(7):30-35.[王钛,许斌,李林国,等.基于离散灰狼算法的多级阈值图像分割[J].计算机技术与发展,2016,26(7):30-35.]
[11]Seth J K,Chandra S.Intrusion detection based on key feature selection using binary GWO[C]//Proceeding of International Conference on Computing for Sustainable Global Development.Piscataway,NJ:IEEE Press,2016:3735-3740.
[12]Emary E,Zawbaa H M,Grosan C,et al.Feature subset selection approach by gray-wolf optimization[C]//Proceedings of Afro-European Conference for Industrial Advancement.S witzerland:Springer,Cham:2014:1-13.
[13]陈昌帅,二进制灰狼优化算法的研究与分析[J].信息系统工程,2016,28(7):136-138.
[14]Yang X S,Deb S.Cuckoo search via Levy flights[C]//Proceedings of Nature&Biologically Inspired Computing,2009.Piscataway,NJ:IEEE Press,2009:210-214.
[15]Jiang Y,Liu X,Yan G,et al.Modified binary cuckoo search for feature selection:a hybrid filter-wrapper approach[C]//Proceedings of International Conference on ComputationalIntelligence and Security.Washington DC:IEEE Computer Society,2017:488-191.
[16]Aziz M A E,Hassanien A E.Modified cuckoo search algorithm with rough sets for feature selection[J].Neural Computing&Applications,2018,29(4):925-934.
[17]Raj A M J,Francis F S,Benadit P J,et al.Particle swarm and cuckoo search(PSCS)optimization based feature selection method to improve the web page classification[J].Journal of Advanced Research in Dynamical&Control Systems,2017,9(6):1125-1140.
[18]Xu H,Liu X,Su J.An improved grey wolf optimizer algorithm integrated with cuckoo search[C]//Proceedings of IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems:Technology and Applications.Piscataway,NJ:IEEE,2017:490-493.
[19]Rodrigues D,Pereira L A M,Almeida T N S,et al.BCS:A binary cuckoo search algorithm for feature selection[C]//Proceedings of IEEE International Symposium on Circuits and Systems.Piscataway,NJ:IEEE,2013:465-468.
[20]Canadian Institute for Cyber security.NSL-KDD dataset[EB/OL].(2009-07-08)[2018-07-05].http://www.unb.ca/cic/datasets/nsl.html.