Font Size: a A A

Study Of Alarm Correlation And Fault Localization Based On Data Mining For Automatically Switched Optical Network

Posted on:2019-06-24Degree:MasterType:Thesis
Country:ChinaCandidate:M C LanFull Text:PDF
GTID:2428330611993320Subject:Engineering
Abstract/Summary:PDF Full Text Request
At present,the automatic switched optical network(ASON)technology has been widely applied in communication field,as the basic transmission platform of information communication,the scale of automatic switched optical network is continuously expanding,and it has proposed higher requirements in optical network operation and maintenance.How to effectively analyze alarm information to realize fault location through clustering correlation,and then enhance reliability of automatic switched optical network communication support is the important task and inevitable trend for operation of automatic switched optical network.In this paper,data mining technology has been applied to alarm correlation and fault localization,and automatic switched optical network alarm correlation and fault location based on data mining is studied.The innovative and practical achievements in this paper mainly include the following aspects:Firstly,the selection of alarm transaction set based on time window directly affects the effectiveness of alarm association rule mining.Meanwhile,the inherent association of intelligent optical network alarm makes it have certain clustering relationship.Therefore,an improved alarm data mining and correlation method is proposed.This method can adapt to the characteristics of internal correlation between multiple planes and multiple network elements of automatic switched optical network,it can effectively compress the quantity of alarm transaction sets and enhance the correctness of alarm correlation rule and provide support for subsequent fault location.Secondly,a fault localization algorithm based on multi-alarm association is proposed.In the guidance of the health management theory,this method uses the large number theorem to weight the alarm,and then it conducts analysis on reasons of multiple alarm combinations,which solves the problem of difficulty in describing network fault correctly of single root alarm and difficulty in fault reason analysis under the situation of multiple alarm combinations,it enhances effectiveness of fault reason location and provides support for subsequent fault diagnosis.Thirdly,this paper designs a prototype system of automatic switched optical network alarm correlation and fault location.This system preliminarily realizes automatic processing of network alarm analysis and fault reason location,assists the network operation and maintenance personnel to diagnose and recover the fault,and provides effective tool for automatic switched optical network operation and maintenance management.
Keywords/Search Tags:Automatic Switched Optical Network, Data Mining, Alarm Correlation, Fault Location
PDF Full Text Request
Related items