Research and Implementation of Distributed Intelligent Network Fault Management System

摘要

The functions of the available network fault management system are imperfect, with relatively poor comprehensiveness, flexibility and intelligence. This paper designs the dynamic distributed architecture of network fault management system, evenly assigning fault management tasks to each domain manager by BP neural network algorithm. In order to intelligently diagnose the fault events as quickly as possible, firstly the system filters the events based on the fault event association tree, then the faults are diagnosed by Case-based Reasoning and CP neural network, getting the conclusion of the faults and adopting some schemes to repair the faults, improving the effect of the fault diagnosis. Finally, a prototype system of network fault management is established and tested. The results show that the system is efficient and intelligent.