Optical Channel Spectrum Anomaly Detection via Trend Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In optical communication systems, existing anomaly detection techniques struggle to accurately identify anomalies in DWDM optical channel spectra due to spectrum tilting and dynamic trends, leading to poor channel discovery and amplifier gain regulation, especially when the number of channels is arbitrary and not on a fixed grid.
Innovation Solution
The method involves obtaining optical channel spectrum data, fitting amplified spontaneous emission (ASE) and channel trends, jointly optimizing these trends to determine an optimized channel trend, and identifying anomalies based on the optimized trend, using robust fitting and unsupervised learning to account for spectrum modifications and tilts without supervision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing anomaly detection techniques are used, then the detection process is simple, but the detection accuracy is poor due to spectrum tilting and dynamic trends
Solution Approach 1:
The patent applies preliminary action by fitting trends to the optical channel spectrum data before performing anomaly detection. The method fits an amplified spontaneous emission trend and a channel trend to the spectrum data, then uses these pre-established trends as references for subsequent anomaly identification. This preliminary trend fitting prepares the data in advance, enabling more accurate anomaly detection despite the added complexity of the fitting process.
2Adaptability or versatility
If fixed thresholds are used for anomaly detection, then the detection method is simple, but it fails to accurately identify anomalies in dynamic spectrum conditions
Solution Approach 1:
The patent applies dynamics by replacing fixed thresholds with dynamic, data-driven thresholds derived from unsupervised learning. The method uses the fitted channel trend to dynamically determine what constitutes normal spectrum variations versus actual anomalies. This allows the system to adapt to changing spectrum conditions, including spectrum tilting and dynamic trends, by continuously adjusting the reference baseline rather than relying on static threshold values.
3Measurement precision
If control channel data is used for anomaly detection, then the detection accuracy may be improved, but the system requires additional control channels and infrastructure
Solution Approach 1:
The patent applies self-service by enabling the optical channel spectrum monitor to perform anomaly detection using only the data from the optical channels themselves, without requiring separate control channels. The unsupervised learning algorithm extracts trends and identifies anomalies directly from the channel spectrum data, allowing the system to be self-sufficient and eliminating the need for additional control channel infrastructure while maintaining accurate channel power estimation and anomaly detection capabilities.
Data Source
AI summary
In some aspects, the techniques described herein relate to a method that includes: obtaining optical channel spectrum data that includes amplified spontaneous emission data and channel data associated with optical signals propagated through an optical fiber; fitting an amplified spontaneous emission trend to the amplified spontaneous emission data; fitting a channel trend to the channel data; jointly optimizing the amplified spontaneous emission trend and the channel trend to determine an optimized channel trend; and determining an anomaly in the channel data based upon the optimized channel trend.