A method and system for optical channel fault diagnosis based on migration learning

Abstract

The invention discloses a method for diagnosing optical channel faults based on migration learning: acquiring optical network performance, alarms, logs and topology data in a certain training area, constructing a data sample set required for model training; extracting the training state from the data sample set Samples, training state features, and training state relationships; select the transfer learning method, input the extracted training state samples, training state features, and training state relationships for model training, and obtain the optical channel state diagnosis training state model; select the optical network performance of the inference area , alarms, logs, and topology data, mark the health status of the optical channel data as inference state samples; load the inference state samples into the optical channel state diagnosis training state model, and train the optical channel state diagnosis inference state model; call the newly generated The optical channel state diagnosis infers the state model to obtain the health status of the analyzed target optical channel. The invention also provides a corresponding optical channel fault diagnosis system based on migration learning.

Description

technical field [0001] The present invention relates to the technical field of OTN equipment management, and more specifically, to a method and system for diagnosing optical channel faults based on migration learning. Background technique [0002] With the rise of cloud computing and 5G interconnection, the demand for network capacity is increasing, and the traditional 10G network is gradually being replaced by 100G. With the rise of data centers, large-scale deployment of 100G backbone networks, more and more switches and routers with 100G Optical Transport Network (OTN) ports, due to the high cost, large size and power consumption of 100G backbone network equipment High, maintenance is difficult, and link failure may directly cause service interruption and affect user experience. Therefore, it is particularly important to predict the sub-health of the optical channel and trace the source of the fault point. [0003] Under the background that different operators have diffe...

AI Technical Summary

Problems solved by technology

There are problems in the existing technology: a single machine learning model is difficult to apply to the OTN network with frequent changes in reality, and transfer learning needs to be introduced to enhance the versatility and universality of traditional machine learning models

Images