基于模糊规则矩阵变换的不确定推理算法研究
|
摘要
本文对传统防火墙技术的的特点进行了分析,并指出其存在的问题,提出研究智能防火墙的必要性,并介绍了当前国内外学者对智能防火墙技术的研究现状。然后根据当前网络信息的不确定性以及不确定推理方法的研究现状,提出研究智能防火墙不确定性推理算法的必要性。
推理以知识为基础,推理系统因拥有知识而具有智能。为了使计算机可以模仿人类的思维行事,就需要把知识用适当的方法表示出来,选用合适的数据结构将它存入计算机,成为计算机可以使用的一种数据。本文介绍了多种知识表示的方法和以知识源分类为基础的知识获取方法,并列举出本文中智能防火墙不确定推理所需要的规则集。
本文概述了目前常用的确定的和不确定的推理方法,并重点对要改进的算法进行了详细的分析论证,发现该算法在环形结构的规则集合和全连通结构规则集中推理效果很好,但在其他拓扑结构中推理能力有限。本文提出在该算法中加入一个相关性分析模块,找出规则集中的条件之间隐含的关系,并将这种关系表示成为模糊产生式规则的形式加入到规则集中,再进行推理。
最后用防火墙规则对改进前后两个算法进行验证,对两个算法的推理结果进行比较后发现,改进后的算法推理能力比原算法要好。
This paper analyzes the characteristics and shortcomings of the technology of traditional firewall, and notes that it's necessary to study the technology of intelligent firewall. Moreover, the research of the scholars at home and abroad on Intelligent Firewall technology was introduced in the paper. Then it suggests that this is necessary to study the inexact reasoning algorithm of intelligent firewall based on the uncertainty information on the network in current and the present situation of uncertainty reasoning method.
Knowledge is the basis of a reasoning system, the reasoning system has intelligence due to it contains knowledge. In order to make the computer simulate human's intelligent action, we need to model the knowledge in an appropriate way and store it into the computer, and then it becomes to be a data structure which the computer can accept. It contains several methods of knowledge representation and the method of knowledge acquisition based on classification of knowledge source in the article. In addition, it enumerates the rule set of intelligent firewall in the paper, which is used in the inexact reasoning.
This article outlines some reasoning methods which were often used currently, including both the exact and the inexact reasoning algorithms. We make a detailed analysis and demonstration on the inexact algorithm which we would like to improve. And we find that the result of reasoning is well in the case that the structure of rules is ring or full connected, but in other topology, like topology tree and topology star, the results of reasoning are not satisfactory. In the paper, we improve the algorithm by adding a relevant analysis module into the old algorithm. The relevant analysis module is used to identify the implicit relationship between the conditions among the rule set, and add this relationship expressed as the form of fuzzy production rules into the rule set, and then perform the reasoning.
Finally, using the firewall rules to verify the reasoning of both the the original and the improved algorithm, we can find that the reasoning ability of improved algorithm is better than the original algorithm by comparing the two results.
推理以知识为基础,推理系统因拥有知识而具有智能。为了使计算机可以模仿人类的思维行事,就需要把知识用适当的方法表示出来,选用合适的数据结构将它存入计算机,成为计算机可以使用的一种数据。本文介绍了多种知识表示的方法和以知识源分类为基础的知识获取方法,并列举出本文中智能防火墙不确定推理所需要的规则集。
本文概述了目前常用的确定的和不确定的推理方法,并重点对要改进的算法进行了详细的分析论证,发现该算法在环形结构的规则集合和全连通结构规则集中推理效果很好,但在其他拓扑结构中推理能力有限。本文提出在该算法中加入一个相关性分析模块,找出规则集中的条件之间隐含的关系,并将这种关系表示成为模糊产生式规则的形式加入到规则集中,再进行推理。
最后用防火墙规则对改进前后两个算法进行验证,对两个算法的推理结果进行比较后发现,改进后的算法推理能力比原算法要好。
This paper analyzes the characteristics and shortcomings of the technology of traditional firewall, and notes that it's necessary to study the technology of intelligent firewall. Moreover, the research of the scholars at home and abroad on Intelligent Firewall technology was introduced in the paper. Then it suggests that this is necessary to study the inexact reasoning algorithm of intelligent firewall based on the uncertainty information on the network in current and the present situation of uncertainty reasoning method.
Knowledge is the basis of a reasoning system, the reasoning system has intelligence due to it contains knowledge. In order to make the computer simulate human's intelligent action, we need to model the knowledge in an appropriate way and store it into the computer, and then it becomes to be a data structure which the computer can accept. It contains several methods of knowledge representation and the method of knowledge acquisition based on classification of knowledge source in the article. In addition, it enumerates the rule set of intelligent firewall in the paper, which is used in the inexact reasoning.
This article outlines some reasoning methods which were often used currently, including both the exact and the inexact reasoning algorithms. We make a detailed analysis and demonstration on the inexact algorithm which we would like to improve. And we find that the result of reasoning is well in the case that the structure of rules is ring or full connected, but in other topology, like topology tree and topology star, the results of reasoning are not satisfactory. In the paper, we improve the algorithm by adding a relevant analysis module into the old algorithm. The relevant analysis module is used to identify the implicit relationship between the conditions among the rule set, and add this relationship expressed as the form of fuzzy production rules into the rule set, and then perform the reasoning.
Finally, using the firewall rules to verify the reasoning of both the the original and the improved algorithm, we can find that the reasoning ability of improved algorithm is better than the original algorithm by comparing the two results.
引文
[1]贾维弟.不确定推理方法研究.哈尔滨工业大学硕士论文,2007.
[2]陈翔.专家系统中不确定推理的研究与应用.安徽大学硕士论文,2006.
[3]周中华.一些不确定性推理方法的研究.江西师范大学硕士论文,2008.
[4]WELLMAN M P. Fundamental concepts of qualitative probabilistic net-works[J]. Artificial Intelligence,1990,44:257～303.
[5]WELLMAN M P. Graphical inference in qualitative probabilistic net works [J]. Net works,1990,20:687～701.
[6]DRUZ DZEL M J, HENR I ON M. Intercausal reasoning with uninstantiated ancestor nodes [C]//Proc of the 11 th Conf UAI,1993:317～325.
[7]Qiansheng Zhang等. An interval-valued fuzzy reasoning algorithm based on OWA operator.IEEE,2009,246～249.
[8]Shyi-Ming Chen, Jyh-Sheng Ke and Jin-Fu Chang. An inexact reasoning algorithm based on fuzzy rule matrix transformation. Journal of Automated Reasoning,1992,8:77-90.
[9]Shyi-Ming Chen, Jyh-Sheng Ke and Jin-Fu Chang. An inexact reasoning algorithm for dealing with inexact knowledge. IJSEKE. Volume:1, Issue:3(1991) pp.227-244.
[10]Chen, S. M. An inexact reasoning technique using linguistic rule matrix transformations. ltiple-Valued Logic,1993.,Page(s):190-195.
[11]Ai-xia Chen, Ning Li and Li Zhao. An inexact reasoning algorithm based on interaction with fuzzy rule matrix transformation. Proceedings of the Eighth International Conference on Machine Learning and Cybernetics, Baoding, July 2009,12-15.
[12]王熙照.模糊测度和模糊积分及在分类技术中的应用.第一版.北京：科学出版社,2008.
[13]黄洪,任卫红等.基于故障树的等级测评专家系统模型研究.计算机应用研究,Vo1.27No.1,Jan,2010.
[14]程传慧,郑秋华等.一种基于二层贝叶斯网的网络人侵检测方法.武汉理工大学学报,Vo1.34 No.1,Feb,2010.
[15]胡家元,曹顺安等.不确定推理在火电厂水汽化学过程故障诊断中的应用.化工自动化及仪表,Vo1.36 No.2,54-58,2009.
[16]王崛,龚长红.装甲车辆故障诊断专家系统.四川兵工学报,Vo1.31 No.2,2010.
[17]孙秋野,张化光等.基于不确定推理的粗糙集配网故障诊断方法研究.自然科学进展,Vo1.19 No.1,2009.
[18]何峻.自动目标识别评估方法研究.国防科技大学博士论文.2009.
[19]张帆,卢昱.基于可信度的不确定推理辅助定密.通信市场,2009.
[20]吴向群.不确定推理在医学诊断预处理系统中的应用.中山大学硕士论文.2006.
[21]Firewalling technology based on the stateful packet filtering. Source:Control & Automation, n 3, p 197-198,2005.
[22]Wes Noonan, Ido Dubrawsky(美).防火墙基础.北京：人民邮电出版社,2007.
[23]Keith E. Strassberg, Richard J. Gondek, Gary Rollie(美).防火墙技术大全北京：机械工业出版,,2003.
[24]Ulrich Ultes-Nitsche, InSeon Yoo. An Intelligent Firewall to Detect Novel Attacks-An Integrated Approach based on Anomaly Detection Against Virus Attacks. Maria Bielikova (Ed.):SOFSEM 2002 Student Research Forum.2002,59～64.
[25]Ulrich Ultes-Nitsche, InSeon Yoo. An integrated network security approach pairing detecting malicious patterns with anomaly detection.
[26]Susan C.Lee, David V.Heinbuch. Training a neural network-based System. IEEE Transactions on Systems, Man, and Cybernetics Part A:Systems and Humans, July, 2001,31(4):294～299.
[27]Miyoung Kim,Misun Kim,Hyewon K.Lee.Youngsong Mun:Design of Active HoneyPot System.ICCSA(2).2003:356～364.
[28]Miyoung Kim,Misun Kim.Youngsong Mun:Design and Implementation of the HoneyPot System with Focusing on the Session Redirection.ICCSA(l).2004:262～269.
[29]Jun Zou, Kaining Lu, Zhigang Jin. Architecture and fuzzy adaptive security algorithm in intellgent firewall. IEEE. British Crown:1145～1149,2002.
[30]吴克喜,赵勤燕,章仁龙,白英彩.具有智能特征的防火墙——专家系统及其在网络安全管理中的应用./小型微型计算机系统.1999,446(6)：446～450.
[31]黄海明,赵海波,杨宇航.“黑洞式”智能型防火墙的关键技术.上海交通大学学报.1998,32(10)：21～25.
[32]许强,江早,赵宏.基于图像内容过滤的智能防火墙系统的研究与实现.计算机研究与发展.2000(4)：221～230.
[33]鄢红国.基于被屏蔽子网混合型智能防火墙设计与实现.武汉理工大学学报,2001,23(4)：54～57
[34]梁柱,曾绍伟.知识表示技术研究.决策管理.2010,No1,52.
[35]年志刚,梁式,麻芳兰,李尚平.知识表示方法研究与应用.计算机应用研究,2007,Vo.124,No.5.
[36]张攀,王波,卿晓霞.专家系统中多种知识表示方法的集成应用.徽型电脑应用, 2004,20(6),4-6
[37]徐舒.产生式表示的改进(硕士学位论文)[D].上海：复旦大学,2005.
[38]张荣沂.专家系统中不确定性知识的表示和处理.计算机应用,2002,21(5),35～39.
[39]林培光,徐如志,余正涛.基于本体和云理论的不确定知识表示.计算机工程与应用,2008,44(5),51～54.
[40]刘思峰,J.福雷斯特.不确定性系统与模型精细化误区.第19届灰色系统全国会议论文,2010.
[41]赵克勤.集对分析及其初步应用[M].浙江科学技术出版社,2000.
[42]安利平,张建勇,仝凌云.专家系统中基于粗集的知识获取、更新与推理.计算机工程与设计,2004,Vo1.25,No.1
[43]韦卫星,莫赞,廖一奎.一种基于神经网络的知识获取方法研究与应用.计算机工程与应用,2004(5),95～98
[44]韦于莉.知识获取研究.情报杂志,2004(4),41～43.
[45]兰美辉,毛雁明等.知识获取研究.电脑知识与技术,2008,Vo1.3,No.9,pp.2056～2057.
[44]胡明晖.基于知识源的专家系统知识获取技术[J].成组技术与生产现代化,2005,22(2)：25～28.
[47]邱志刚.智能防火墙过滤规则学习与优化.哈尔滨工业大学硕士论文,2006.5：27～28
[48]刘伟娜,霍利民,张立国.贝叶斯网络确定推理算法的研究.微计算机信息,2006,Vol 22(3),92～94.
[49]刘俊娜,张佑生.一种自适应的联合树算法.系统仿真技术及其应用,2006(8),152～155.
[50]严慧鑫.贝叶斯网确定推理算法的研究.吉林大学硕士论文.2005.
[51]Stuart Russell,Peter Norvig.Artificial intelligence:a modern approach [M].Englewood Cliffs, NJ:Prentice Hall,1995.
[52]Adnan Darwiche. An introduction to inference in Bayesian networks.Number D-102.UCLA Cognitive Systems Laboratory, May,1999.
[53]Rose, D., Tarjan, R., and Lueker, G.Algorithmic aspects of vertex elimination on graphs.SIAM Journal of Computing,5(2),1976:266-283.
[54]Nevin LianWen Zhang, David Poole.A simple approach to Bayesian network computations.Proceedings of the 10th Canadian Conference on Artificial Intelligence, 1994,171～178.
[55]孙寒星.雷达辐射源识别专家系统中推理机的开发与研究.解放军信息工程大学硕士 论文,2005.
[56]Shaefr G. A Mathematieal Theory of Evidence.Princeton, N J:Princeton UP,1976, 10～40
[2]陈翔.专家系统中不确定推理的研究与应用.安徽大学硕士论文,2006.
[3]周中华.一些不确定性推理方法的研究.江西师范大学硕士论文,2008.
[4]WELLMAN M P. Fundamental concepts of qualitative probabilistic net-works[J]. Artificial Intelligence,1990,44:257～303.
[5]WELLMAN M P. Graphical inference in qualitative probabilistic net works [J]. Net works,1990,20:687～701.
[6]DRUZ DZEL M J, HENR I ON M. Intercausal reasoning with uninstantiated ancestor nodes [C]//Proc of the 11 th Conf UAI,1993:317～325.
[7]Qiansheng Zhang等. An interval-valued fuzzy reasoning algorithm based on OWA operator.IEEE,2009,246～249.
[8]Shyi-Ming Chen, Jyh-Sheng Ke and Jin-Fu Chang. An inexact reasoning algorithm based on fuzzy rule matrix transformation. Journal of Automated Reasoning,1992,8:77-90.
[9]Shyi-Ming Chen, Jyh-Sheng Ke and Jin-Fu Chang. An inexact reasoning algorithm for dealing with inexact knowledge. IJSEKE. Volume:1, Issue:3(1991) pp.227-244.
[10]Chen, S. M. An inexact reasoning technique using linguistic rule matrix transformations. ltiple-Valued Logic,1993.,Page(s):190-195.
[11]Ai-xia Chen, Ning Li and Li Zhao. An inexact reasoning algorithm based on interaction with fuzzy rule matrix transformation. Proceedings of the Eighth International Conference on Machine Learning and Cybernetics, Baoding, July 2009,12-15.
[12]王熙照.模糊测度和模糊积分及在分类技术中的应用.第一版.北京：科学出版社,2008.
[13]黄洪,任卫红等.基于故障树的等级测评专家系统模型研究.计算机应用研究,Vo1.27No.1,Jan,2010.
[14]程传慧,郑秋华等.一种基于二层贝叶斯网的网络人侵检测方法.武汉理工大学学报,Vo1.34 No.1,Feb,2010.
[15]胡家元,曹顺安等.不确定推理在火电厂水汽化学过程故障诊断中的应用.化工自动化及仪表,Vo1.36 No.2,54-58,2009.
[16]王崛,龚长红.装甲车辆故障诊断专家系统.四川兵工学报,Vo1.31 No.2,2010.
[17]孙秋野,张化光等.基于不确定推理的粗糙集配网故障诊断方法研究.自然科学进展,Vo1.19 No.1,2009.
[18]何峻.自动目标识别评估方法研究.国防科技大学博士论文.2009.
[19]张帆,卢昱.基于可信度的不确定推理辅助定密.通信市场,2009.
[20]吴向群.不确定推理在医学诊断预处理系统中的应用.中山大学硕士论文.2006.
[21]Firewalling technology based on the stateful packet filtering. Source:Control & Automation, n 3, p 197-198,2005.
[22]Wes Noonan, Ido Dubrawsky(美).防火墙基础.北京：人民邮电出版社,2007.
[23]Keith E. Strassberg, Richard J. Gondek, Gary Rollie(美).防火墙技术大全北京：机械工业出版,,2003.
[24]Ulrich Ultes-Nitsche, InSeon Yoo. An Intelligent Firewall to Detect Novel Attacks-An Integrated Approach based on Anomaly Detection Against Virus Attacks. Maria Bielikova (Ed.):SOFSEM 2002 Student Research Forum.2002,59～64.
[25]Ulrich Ultes-Nitsche, InSeon Yoo. An integrated network security approach pairing detecting malicious patterns with anomaly detection.
[26]Susan C.Lee, David V.Heinbuch. Training a neural network-based System. IEEE Transactions on Systems, Man, and Cybernetics Part A:Systems and Humans, July, 2001,31(4):294～299.
[27]Miyoung Kim,Misun Kim,Hyewon K.Lee.Youngsong Mun:Design of Active HoneyPot System.ICCSA(2).2003:356～364.
[28]Miyoung Kim,Misun Kim.Youngsong Mun:Design and Implementation of the HoneyPot System with Focusing on the Session Redirection.ICCSA(l).2004:262～269.
[29]Jun Zou, Kaining Lu, Zhigang Jin. Architecture and fuzzy adaptive security algorithm in intellgent firewall. IEEE. British Crown:1145～1149,2002.
[30]吴克喜,赵勤燕,章仁龙,白英彩.具有智能特征的防火墙——专家系统及其在网络安全管理中的应用./小型微型计算机系统.1999,446(6)：446～450.
[31]黄海明,赵海波,杨宇航.“黑洞式”智能型防火墙的关键技术.上海交通大学学报.1998,32(10)：21～25.
[32]许强,江早,赵宏.基于图像内容过滤的智能防火墙系统的研究与实现.计算机研究与发展.2000(4)：221～230.
[33]鄢红国.基于被屏蔽子网混合型智能防火墙设计与实现.武汉理工大学学报,2001,23(4)：54～57
[34]梁柱,曾绍伟.知识表示技术研究.决策管理.2010,No1,52.
[35]年志刚,梁式,麻芳兰,李尚平.知识表示方法研究与应用.计算机应用研究,2007,Vo.124,No.5.
[36]张攀,王波,卿晓霞.专家系统中多种知识表示方法的集成应用.徽型电脑应用, 2004,20(6),4-6
[37]徐舒.产生式表示的改进(硕士学位论文)[D].上海：复旦大学,2005.
[38]张荣沂.专家系统中不确定性知识的表示和处理.计算机应用,2002,21(5),35～39.
[39]林培光,徐如志,余正涛.基于本体和云理论的不确定知识表示.计算机工程与应用,2008,44(5),51～54.
[40]刘思峰,J.福雷斯特.不确定性系统与模型精细化误区.第19届灰色系统全国会议论文,2010.
[41]赵克勤.集对分析及其初步应用[M].浙江科学技术出版社,2000.
[42]安利平,张建勇,仝凌云.专家系统中基于粗集的知识获取、更新与推理.计算机工程与设计,2004,Vo1.25,No.1
[43]韦卫星,莫赞,廖一奎.一种基于神经网络的知识获取方法研究与应用.计算机工程与应用,2004(5),95～98
[44]韦于莉.知识获取研究.情报杂志,2004(4),41～43.
[45]兰美辉,毛雁明等.知识获取研究.电脑知识与技术,2008,Vo1.3,No.9,pp.2056～2057.
[44]胡明晖.基于知识源的专家系统知识获取技术[J].成组技术与生产现代化,2005,22(2)：25～28.
[47]邱志刚.智能防火墙过滤规则学习与优化.哈尔滨工业大学硕士论文,2006.5：27～28
[48]刘伟娜,霍利民,张立国.贝叶斯网络确定推理算法的研究.微计算机信息,2006,Vol 22(3),92～94.
[49]刘俊娜,张佑生.一种自适应的联合树算法.系统仿真技术及其应用,2006(8),152～155.
[50]严慧鑫.贝叶斯网确定推理算法的研究.吉林大学硕士论文.2005.
[51]Stuart Russell,Peter Norvig.Artificial intelligence:a modern approach [M].Englewood Cliffs, NJ:Prentice Hall,1995.
[52]Adnan Darwiche. An introduction to inference in Bayesian networks.Number D-102.UCLA Cognitive Systems Laboratory, May,1999.
[53]Rose, D., Tarjan, R., and Lueker, G.Algorithmic aspects of vertex elimination on graphs.SIAM Journal of Computing,5(2),1976:266-283.
[54]Nevin LianWen Zhang, David Poole.A simple approach to Bayesian network computations.Proceedings of the 10th Canadian Conference on Artificial Intelligence, 1994,171～178.
[55]孙寒星.雷达辐射源识别专家系统中推理机的开发与研究.解放军信息工程大学硕士 论文,2005.
[56]Shaefr G. A Mathematieal Theory of Evidence.Princeton, N J:Princeton UP,1976, 10～40
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |