Multiple Fault Identification using Neural Models for Telecommunication Networks

A. Yilmaz and İ. Kılınç (Turkey)


Telecommunications network, network management, neural network, alarm correlation, fault identification, coding, minimal distance decoder


Alarm management is one of the basic management functions in modern telecommunications network including ATM networks. Any problem occurring in the network can generate single or multiple alarms defined by the nature of interconnection. Artificial intelligence methods and their applications exist and are used in the alarm lifecycle especially for alarm handling. One of the correlation techniques used in this area is the coding approach. In this paper we analyzed some scenarios and present the advantages of the artificial neural network (NN) over minimal distance decoder to increase the ability of identifying multiple faults that might happen simultaneously.

Important Links:

Go Back