电信网络中的告警关联与通话管理
|
摘要
本文结合两个与企业合作的工程项目:“天津移动告警关联分析系统(TACAS)”和某运营商的NGN业务网络中的“通话管理能力集”,对告警关联挖掘技术和Webservice分布式技术在网络管理系统中的应用及其实现方案进行了研究。
天津移动告警关联分析系统,利用电信领域数据挖掘的理论知识,研发了一系列的关联规则挖掘算法,本文介绍了常用的关联规则挖掘算法,深入分析了TACAS系统中应用的WINEPI串行情节挖掘算法,并对原算法做了修改,提出采用相差度替换频繁度的DWinepi算法,该算法有效的减少了挖掘过程中产生的频繁集数目,提高了算法效率,而且挖掘出来的结果也更为严谨。此外,本文完整介绍了TACAS系统数据挖掘流程的各个阶段,包括告警的导入、预处理、告警关联挖掘以及挖掘结果的评估和展现。
现有的电信网络管理系统一般由许多独立管理的专业网管系统构成,大多采用各自的通信协议,互不兼容,网络管理系统由一些互通性较差的多个管理域组成,给管理和业务发展带来很多不便;本文结合《通话管理能力集》的设计实现方案,把Webservice分布式技术引入到网络管理能力集中,把已有的网络管理功能封装起来,通过封装功能提高网络管理系统的扩展性和灵活性,在网络异构环境下跨越防火墙和各种平台实现组件的互操作。此外,对通话管理能力集的整体架构和流程,本文也进行了描述。
Based on the two projects in the cooperation with enterprises: " China Mobile Tianjin Branch Alarm association Analysis System" and "Communicatin Management Group" which is in NGN business network,the paper focus on the application of alarm association data mining and Webservice in the network management system.
In China Mobile Tianjin Branch Alarm association Analysis System , a series of data mining algorithm is developed.Firstly,typical algorithm of alarm association rules is presented and then it analyzes the serial algorithom of WINEPI applied in the system. Furthermore it bring forward an improved algorithm called DWinepi.In DWinepi,it introduces discrepancy instead of frequency and frequent episodes can be reduced effectively by this method. So DWinepi is more effective and the rules mined by it will be more precise . In addition, the paper completely analyses each phase of the data mining process, including the data importing, pretreatment, assciation rules excavation and the results assessment and display.
With the development of the telecom-networks, the quantity and kinds of equipments keep increasing and the structure become complex. However, the telecom-network management system is composed of many separated special subsystems at present, which communicate with different protocols. This causes that each subsystem, serving special sub-network, can not be compatible with others.Nonstandard multi-field of management result that many management fields with bad exchange ability build up the management system. The management system can not keep pace with the requirement of the telecommunication services providers.Based on the project cooperate with entiprise:"Communication Management Capability Group",this paper focus on the application of Webservice in network management,which is cross-platform, cross-language and loose-coupling.Existed function of network management can be encapsulated via Webservice and this can improve the expansibility .
天津移动告警关联分析系统,利用电信领域数据挖掘的理论知识,研发了一系列的关联规则挖掘算法,本文介绍了常用的关联规则挖掘算法,深入分析了TACAS系统中应用的WINEPI串行情节挖掘算法,并对原算法做了修改,提出采用相差度替换频繁度的DWinepi算法,该算法有效的减少了挖掘过程中产生的频繁集数目,提高了算法效率,而且挖掘出来的结果也更为严谨。此外,本文完整介绍了TACAS系统数据挖掘流程的各个阶段,包括告警的导入、预处理、告警关联挖掘以及挖掘结果的评估和展现。
现有的电信网络管理系统一般由许多独立管理的专业网管系统构成,大多采用各自的通信协议,互不兼容,网络管理系统由一些互通性较差的多个管理域组成,给管理和业务发展带来很多不便;本文结合《通话管理能力集》的设计实现方案,把Webservice分布式技术引入到网络管理能力集中,把已有的网络管理功能封装起来,通过封装功能提高网络管理系统的扩展性和灵活性,在网络异构环境下跨越防火墙和各种平台实现组件的互操作。此外,对通话管理能力集的整体架构和流程,本文也进行了描述。
Based on the two projects in the cooperation with enterprises: " China Mobile Tianjin Branch Alarm association Analysis System" and "Communicatin Management Group" which is in NGN business network,the paper focus on the application of alarm association data mining and Webservice in the network management system.
In China Mobile Tianjin Branch Alarm association Analysis System , a series of data mining algorithm is developed.Firstly,typical algorithm of alarm association rules is presented and then it analyzes the serial algorithom of WINEPI applied in the system. Furthermore it bring forward an improved algorithm called DWinepi.In DWinepi,it introduces discrepancy instead of frequency and frequent episodes can be reduced effectively by this method. So DWinepi is more effective and the rules mined by it will be more precise . In addition, the paper completely analyses each phase of the data mining process, including the data importing, pretreatment, assciation rules excavation and the results assessment and display.
With the development of the telecom-networks, the quantity and kinds of equipments keep increasing and the structure become complex. However, the telecom-network management system is composed of many separated special subsystems at present, which communicate with different protocols. This causes that each subsystem, serving special sub-network, can not be compatible with others.Nonstandard multi-field of management result that many management fields with bad exchange ability build up the management system. The management system can not keep pace with the requirement of the telecommunication services providers.Based on the project cooperate with entiprise:"Communication Management Capability Group",this paper focus on the application of Webservice in network management,which is cross-platform, cross-language and loose-coupling.Existed function of network management can be encapsulated via Webservice and this can improve the expansibility .
引文
[1]郭军,网络管理(第2版),北京邮电大学出版社,2006.4
[2]毛国军,段立娟,数据挖掘原理与算法,清华大学出版社编著,2005.7,65-70
[3]Mack Hendricks著卢刚,高春荣,朱军译《Java Web服务编程指南》,电子工业出版社,2002.10,20-50
[4]Boutaba,R,Go lab,W.Iraqi,Y,Light paths on demand:a Web-services-based management system[J]Communications Magazine,IEEE,2004,42(7):101-107
[5]吴扬阳,陈怀南,基于关联规则的通信网络告警相关性分析模型[J],通讯和计算机,2004.12,57-63
[6]徐磊等,基于WebServices的网络管理实现模式的研究,2005.1
[7]HEIKKI MANNILA,HANNU TOIVONEN A,INKERI VERKAMO Discovery of Frequent Episodes in Event Sequences Data Mining and Knowledge Discovery 1,259-289(1997)
[8]Anand Tripathi.Challenges Designing Next Generation Middleware Systems.Communications of the ACM,2002,45(6):38-42.
[9]H.Mannila,H.Toivonen,A.Verkamo.Discovery of frequent episodes in event sequences[J].Data Mining and Knowledge Discovery,1997,1(3):259-289.2001.8
[10]桑蚕,网管系统与应用系统的整合趋势,网管员世界,2006年3期
[11]Deng Yong(邓勇),Liu Qi,Li Yi Yue An efficient algorithm to generate candidates in discovering frequent episodes 上海交通大学 高技术通讯:英文版-2006年1期
[12]丁军,数据挖掘及其在电信告警中的应用,山东大学,2003.11
[13]刘娜,大连网通网管告节过滤机制的研究与应用,大连理工大学,2007.6
[14]肖海林,网络告警关联规则挖掘系统的实现,电子科技大学,2007.5
[15]TACASv5.0详细设计文档 2006.5
[16]TACASv5.0总体方案设计 2006.5
[17]单莘、朱永宣、郭军 电信网告警数据库中的增量式挖掘技术研究 计算机应用研究 2006.3
[18]天津移动网络生产管理中心告警关联分析系统项目工作报告 2006.7
[2]毛国军,段立娟,数据挖掘原理与算法,清华大学出版社编著,2005.7,65-70
[3]Mack Hendricks著卢刚,高春荣,朱军译《Java Web服务编程指南》,电子工业出版社,2002.10,20-50
[4]Boutaba,R,Go lab,W.Iraqi,Y,Light paths on demand:a Web-services-based management system[J]Communications Magazine,IEEE,2004,42(7):101-107
[5]吴扬阳,陈怀南,基于关联规则的通信网络告警相关性分析模型[J],通讯和计算机,2004.12,57-63
[6]徐磊等,基于WebServices的网络管理实现模式的研究,2005.1
[7]HEIKKI MANNILA,HANNU TOIVONEN A,INKERI VERKAMO Discovery of Frequent Episodes in Event Sequences Data Mining and Knowledge Discovery 1,259-289(1997)
[8]Anand Tripathi.Challenges Designing Next Generation Middleware Systems.Communications of the ACM,2002,45(6):38-42.
[9]H.Mannila,H.Toivonen,A.Verkamo.Discovery of frequent episodes in event sequences[J].Data Mining and Knowledge Discovery,1997,1(3):259-289.2001.8
[10]桑蚕,网管系统与应用系统的整合趋势,网管员世界,2006年3期
[11]Deng Yong(邓勇),Liu Qi,Li Yi Yue An efficient algorithm to generate candidates in discovering frequent episodes 上海交通大学 高技术通讯:英文版-2006年1期
[12]丁军,数据挖掘及其在电信告警中的应用,山东大学,2003.11
[13]刘娜,大连网通网管告节过滤机制的研究与应用,大连理工大学,2007.6
[14]肖海林,网络告警关联规则挖掘系统的实现,电子科技大学,2007.5
[15]TACASv5.0详细设计文档 2006.5
[16]TACASv5.0总体方案设计 2006.5
[17]单莘、朱永宣、郭军 电信网告警数据库中的增量式挖掘技术研究 计算机应用研究 2006.3
[18]天津移动网络生产管理中心告警关联分析系统项目工作报告 2006.7