Optical Performance Monitoring Using Artificial Neural Network Trained With Asynchronous Amplitude Histograms

Optical Performance Monitoring Using Artificial Neural Networks Trained With Empirical Moments of Asynchronously Sampled Signal Amplitudes

“…In this letter, we propose the use of ANN trained with empirical moments of asynchronously sampled signal amplitudes for multi-impairment monitoring in 40/56-Gb/s RZ-DQPSK and 40-Gb/s RZ-DPSK systems. Simulation results demonstrate simultaneous and independent OSNR, signed CD, and PMD monitoring with wide dynamic ranges and good monitoring accuracies which are comparable with existing ANN-based techniques [2][3][4][5][6]. Unlike existing ANNbased methods, the proposed technique enables monitoring of both magnitude and sign of accumulated CD.…”

“…The introduction of CD shift has two useful outcomes: (1) The moments µ i,offset are no longer symmetrical for positive and negative CD values and thus enable the discrimination of sign of accumulated CD values. Therefore, unlike existing ANN-based techniques [2][3][4][5][6], which can only monitor the magnitude of accumulated CD, the proposed technique can perform signed CD monitoring. (2) The set of ten empirical moments [µ 1 , µ 2 , ... .…”

“…Therefore, it is imperative to incorporate effective monitoring mechanisms across the whole fiber-optic communication network which could provide precise and real-time information about the health of each individual wavelength-division multiplexed (WDM) channel. Optical performance monitoring (OPM) is indispensable for the reliable operation, cost-effective self-management and efficient utilization of available network resources in emerging dynamic fiber-optic networks, in which individual WDM channels may traverse different paths and hence may Manuscript [3,4], asynchronous constellation diagrams [5], asynchronous amplitude histograms [6] etc. and exploit them for the training of ANN and subsequently multiimpairment monitoring.…”

An Overview on Application of Machine Learning Techniques in Optical Networks

“…Artificial neural networks are well suited machine learning tools to perform optical performance monitoring as they can be used to learn the complex mapping between samples or extracted features from the symbols and optical fiber channel parameters, such as OSNR, PMD, Polarization-dependent loss (PDL), baud rate and CD. The features that are fed into the neural network can be derived using different approaches relying on feature extraction from: 1) the power eye diagrams (e.g., Q-factor, closure, variance, root-meansquare jitter and crossing amplitude, as in [49]- [53], [69]); 2) the two-dimensional eye-diagram and phase portrait [54]; 3) asynchronous constellation diagrams (i.e., vector diagrams also including transitions between symbols [51]); and 4) histograms of the asynchronously sampled signal amplitudes [52], [53]. The advantage of manually providing the features to the algorithm is that the NN can be relatively simple, e.g., consisting of one hidden layer and up to 10 hidden units and does not require large amount of data to be trained.…”

Investigation of OSNR monitoring technique based on the beat noise detection

“…Optical performance monitoring is one of the key technologies for the future ultra-high speed optical networks [1][2][3][4]. Optical signal-to-noise ratio (OSNR) directly reflects the quality of the signal.…”