A deep neural network model construction method, fault diagnosis method and system

Abstract

The invention discloses a construction method of a deep neural network model, a fault diagnosis method and a system, and relates to the technical field of communication. The construction method includes: determining the alarm root derivation rule based on the service path, in the service topology of the target network, the adjacent service nodes in the service path from the source end to the sink end are the relationship between the client layer and the service layer; based on expert fault diagnosis data construction A unified diagnostic factor matrix for diagnosing faults in service paths. The unified diagnostic factor matrix includes the root node and the alarms and performance status indicators on the service nodes associated with the root node. The root node is the service node that generates the root alarm; The matrix is ​​used as input, and the probability vector of fault cause type is used as output to build a deep neural network model, and use sample data for training and validation. The present invention is based on comprehensive and effective alarms and performance state indicators, so that the constructed model can quickly and accurately perform fault diagnosis and location.

Description

technical field [0001] The present invention relates to the technical field of communication, in particular to a method for constructing a deep neural network model, a method for diagnosing faults and a system. Background technique [0002] With the continuous increase of network services brought about by the all-round development of the society, the scale of the network has become larger and the network environment has become more complex, and the types and times of faults have also increased, making the investment of major network operators in troubleshooting. bigger. [0003] In traditional manual troubleshooting, the first step is to obtain key alarms or root cause alarms from a large number of alarms, and then query other related information (such as minor or derivative alarms, performance, status, configuration, etc.) on the service topology link based on these alarms. , and finally determine the cause of the failure and determine the location of the failure. [0004...

AI Technical Summary

Problems solved by technology

However, the current deep learning solutions all have different degrees of defects, and it is difficult to achieve the desired effect when put into the network application.

Images