Design And Implementation Of Alarm Subsystem In IP Network Management System

We proprosed a new IP bearer network alarm management system based on fuzzy timeseries approach by combining alarm prediction technology and IP bearer network. Thestable operation of the alarm management system plays an important role in the normaloperation of the network management system. Its business security and function plays ahuge impact on the investment. With the construction reaches a certain size, the activemonitoring and management of alarm information for IP bearer networks becomes a veryimportant issue which must be considered. With the widely improvement of theperformance of our computers, monitoring real-time operation and alarm status of IP bearernetworks becomes more and more practical issues rather than theoretical studies. How dowe study the convetional stochastic process theory based forecasting methods? How do weuse fuzzy logic in the technical application of the principle of the time series? How do wepredict network performace by analyzing the performance of switches, alarms,configuration and basic data? The research projects related to these questions are verysignificant and have high research value.IP bearer network’s active monitoring is able to find network anomalies, providealarms and take actions (such as SMS notification, processes, optimization strategies andother issues) by predicting technology. This project focuses on the study of activemonitoring technology of IP bearer network. We aim to improve the management, earlywarning and optimize network performance. The active monitoring real-time performancealarm algorithm was used to calculate the "baseline" and "line" of tolerance for predictionand management of alarms in the system. In this paper, the development of alarmmanagement subsystem was described in detail. The alam subsystem is able to generatealams and network event reports by collecting data from the front system, and send thisreport to the network managers. In the meanwhile, the alarm data can be normalized,classified and backed up before they stored in the database. The stored historicalinformation can be queried and analyzed as well. By testing database, processes and upperapplication interfaces, it was proved that the alarm module has proper function, goodperformance, robust structure and meets all the design requirements.This article is composed of two major parts. In the first part we introduced the researchbackground and significance of the project, the current status of this field and relatedtheoretical studies, including IP bearer network theory, the development and current statusof the prediction technology and so on. In the second part, we first briefly described the IP bearer network management system of organization and structure, functions, and then wedescribed the development of alarm management subsystem in more details. In thedescription of the overall design of the network management system, we introduced itsorganization, framework and functions. In the description of the alarm managementsubsystem, we first introduced its design objectives, related concepts and overallframework, and then we described its main function modules in detail, including thedescription of data, message, alarm management, and so on.