Design And Implementation Of The Processing Analysis System Of Alarm Data In Mobile Base Stations

The base station system of 2G and 4G network is responsible for the user access and wireless resource management, has important significance in the whole mobile network. Now the traditional passive reactive processing of base station alarm cannot meet the demand of network fault management for the mobile operator any more, a new network management model with analysis of alarm data and forwardly find of the problem of network is future development direction of alarm fault management mode.This paper design and implement a processing and analysis system of mobile network base station alarm data for one City Company of China Mobile. The system is used to statistics and analyse the alarm data of base station system, give an evaluation value of the whole network, mine the occurrence regular pattern of stop services alarm in historical alarm data, and on this basis, get the reasons of stop services alarm of base station. Thus, it is helpful for network devices maintaining staffs to master the operation state of the network, timely adjust the strategy of equipments maintenance according to the analysis results, and ensure the stable operation of the network equipment. The processing and analysis system of base station alarm data is mainly composed of alarm data statistics module, evaluation of network status modules and reason analysis of stop services alarm module. Alarm data statistics module realizes the present of base station alarm data from different aspects, including alarm data of 2G and 4G networks, the module to help users select base station with stop services take a long time or more frequent, master the informatin about proportion distribution of stop services alarm in the different regional or maintainer. Network evaluation module design a reasonable indicator system, using FAHP evaluation method with expert advices quantize the overall evaluation of mobile network base station system, and using CART algorithm realizes the automatic evaluation, to help users grasp the network status and its changes on the whole. Reason analysis of stop services alarm module relaizesthe automatic reason judgment of stop services base station, the module is divided into mining module of regular pattern of base station stop services alarm and expert system module of rules application. This paper do experiments test the operation efficiency of the mining model based on Apriori algorithm and FP-Growth algorithm, then choose the mining model base on FP-Growth algorithm. In the post-processing of the rules, this paper propose a model based on the principle of Apriori, data constraints and support sorting, to reasonably and effectively show the rules for assessment. Using these association rules by rule-based expert system, the association rules combined with expert experience and konwledge will be converted into rules stored in the knowledge base of expert system. Then design reasoning mechanism of fuzzy matching and control mechanism of data driven to relaize the judgment of stop services reasons of base station.The processing and analysis system of mobile network base station alarm data has been running stably in one City Company of China Mobile, the system is able to present base station alarm data from multiple perspective; According to the monthly alarm data, the quantitative value and change of network state is given by the system; the system can obtain regular patterns of stop services alarm of base station, apply regular patterns by expert systems in practice, and realize reason judgment of stop services alarm, and meet the demand of mobile company business.