Research Of Data Mining And It's Applications In Telecom Network Management System

The constructions of the GSM network of Telecom is complexes and diverse, composed of network-element from many different manufacturers. One of the network-element's failures may be associated with multiple devices and links which may caucus a large number of derivative alarm.These alarm data are collected and deposited to the alarm database by the network management system of telecommunications, and finally presented to the user by alarm management system. The alarm database of Telecom conserves a large number of historical alarm information, which contains much useful information reflecting the regulation of the network's operation. Alarm data is very suitable for the application of the association rule analysis, and the results of application can improve the network failure management.In this paper, we introduce the basic concepts and related knowledge of the data warehouse and data mining, as well as for the practice of data mining methods and steps. Then we introduce the basic concepts and related knowledge and the status quo of the telecom network management systems and the communication monitoring system, elaborate the theory of association rule mining algorithm and the basic method applied for communication network alarm correlation analysis detailed, analysis the related characteristics of alarm data, and compared with the method of the available alarm correlation analysis, then discuss a number of relevant alarm approach.Finally, we introduce the specific realization methods of the analysis of the alarm correlation, and the results of the application in the network management system based on the results of research above.