Development And Application Of Alarm Analysis Platform For Telecom Key Client Network Management

In order to provide a better service for various users,telecom operators around the world are continuously improving their network management capability while improving the data transmission speed.Key client can bring huge benefits to telecom operators,and have always been the focus of telecom operators' development.In order to provide more efficient and accurate services to the key client,telecom operators established the key client network management system.Nowadays,because of the large-scale,complex,and diverse data of the key client network management,it becomes more difficult to analyze the alarm data.The big data technology can improve the analysis efficiency of massive and complex data.Therefore,building a big data analysis platform that can comprehensively analyze the characteristics of network alarms and faults,find faults and hidden dangers in a timely manner,and determine the cause of failure quickly is of great significance to the improvement of fault management capability of key client network management.This thesis is from telecommunication operator's cooperation project.This thesis used big data technology to design and implement an alarm analysis platform for telecom key client network management,and the platform is used to analysis the alarm data.This thesis's work contains the following three aspects:(1)Related technical review.We summarized the status and development of network management system.Then,we summarized the big data technology and alarm analysis technology required for the development and application of the platform.(2)Based on Spark,the alarm analysis platform for telecom key client network management was designed and implemented.This thesis analyzed the needs of the company to complete the overall design of the platform,and divided the platform into five modules:data acquisition,parsing and cleaning,storage,analysis,visualization.Then,this thesis dissertated the design and implementation of each module based on the data characteristics and function of the module.Finally,this thesis tested the performance and functionality of the platform(3)The alarm analysis platform for telecom key client network management was used to analyze the alarm data,which achieved the network management knowledge base,fault location and fault diagnosis application.In this thesis,we first built a network management knowledge base by mining the potential relationship between alarms and faults through data analysis.Then,we built fault localization and fault diagnosis models through neural network technologies,which are used to determine the fault location and fault type respectively.Further,in this thesis,the fault location model and fault diagnosis model are imported into the platform's data analysis module.Then,the platform is used for real-time analysis of alarms to locate and diagnose the faults rapidly by using the models.