Anomaly Detection in Optical Spectra via Joint Optimization

Antonino Maria Rizzo¹, Luca Magri¹, Pietro Invernizzi², Enrico Sozio², Stefano Piciaccia², Alberto Tanzi², Stefano Binetti², Cesare Alippi^1,3, Giacomo Boracchi^1
1Politecnico di Milano, ²Cisco Photonics, ³Università della Svizzera Italiana

Abstract

Despite the remarkable progress of fiber optics in communication, little attention has been devoted to the automatic detection of anomalies in optical spectra, i.e., poorly transmitted channels. This task is typically addressed by ad-hoc heuristics that fall short in spectra presenting heavy distortions caused by optical amplifiers during transmission. We propose a method based on a joint optimization procedure for estimating the major trends that characterize the spectrum, enabling the detection of anomalies even in the presence of few channels and heavy distortions. Our experiments have shown that the proposed method can successfully localize anomalies achieving more than 98% accuracy, outperforming all competitors.

Paper Poster

Introduction

• User information is enclosed into spectra and transmitted through fiber optics.
• A spectrum contains a varying number of channels that follow a trend 𝜈.
• The occurrence of faults causes anomalies.
• Our goal is to automatically detect the anomalies in the spectrum, if any.

Key Intuitions

• Anomalies are channels that do not conform to the channel trend.
• Fit the channel trend via robust fitting, such as RanSaC.
• Joint optimization allows us to promote the similarity between channel and ASE trends.

Joint Optimization

Our method account for the similarity of the channel and ASE trends, 𝜈 and 𝜇, by optimizing the following loss:

Proposed Method

Running Example

Experiments

Our experiments demonstrate that our joint optimization procedure enhances channel trend estimation, enabling the accurate detection of anomalies.