An update method and system for predicting optical performance degradation trend

Abstract

An update method and system for predicting optical performance degradation trend. The present invention relates to the field of optical communication. The method includes the steps of: constructing a neural network for predicting optical performance degradation trend based on original key performance indicators; extracting new key points from the increased optical performance indicators The performance index is obtained by calculating the covariance between the new key performance index and the original key performance index to obtain the correlation coefficient, and the new key performance index whose correlation coefficient is less than or equal to the influence degree threshold is selected as all or part of the samples to update the neural network. When the optical performance index increases, the present invention obtains a new key performance index, and uses it as a sample to update the neural network for predicting optical performance degradation, continuously improves the optical performance degradation trend, and improves the accuracy of prediction.

Description

technical field [0001] The present invention relates to the field of optical communication, in particular to an update method and system for predicting the degradation trend of optical performance. Background technique [0002] During the operation of optical communication equipment, optical devices, optical modules, and optical channels continuously generate a large amount of performance status data, which are performance indicators; these data are obtained by the optical module and single disk of the equipment, and reported to the management and control platform for processing. [0003] Traditional optical performance degradation prediction methods include Markov transfer matrix prediction, exponential smoothing prediction, etc. These methods mine the variation law of optical performance indicators, and predict the degradation trend of optical performance indicators based on the mined variation laws. [0004] However, these methods need to obtain the key performance indica...

AI Technical Summary

Problems solved by technology

[0004] However, these methods need to obtain the key performance indicators that affect the prediction of the deterioration trend. If some key performance indicators cannot be obtained, the prediction error will be relatively large

In actual communication network engineering, affected by various factors, some key performance indicators may not be obtained in time

For example, communication interruptions lead to failure to report optical power performance indicators, optical module manufacturers are temporarily unable to provide key performance indicators such as dispersion, and management and control platform hard disk failures lead to loss of optical signal-to-noise ratio performance indicators. The above situations will lead to some key optical performance indicators cannot be extracted. Affect the forecasting effect of traditional forecasting methods

Images