Optical Signal-to-noise Ratio Monitoring Research Articles

Transfer learning-assisted modulation classification and OSNR estimation using constellation diagrams

Abstract A cost-effective deep transfer learning (TL) learned with characteristics derived from the constellation diagram is proposed for executing modulation classification and optical signal-to-noise ratio (OSNR) estimation for the future generation of elastic optical networks. The information acquired with the ImageNet dataset is transferred using pre-trained TL versions including VGG19, ResNet152V2, InceptionV3, and ResNet50 to identify various modulation formats and their corresponding OSNR. Adam optimizer is employed to ascertain the appropriate settings for the hyperparameters. Numerical results demonstrate that the proposed VGG19 method achieves the highest level of accuracy (100%) in the recognition of various modulation forms. To fulfil the requirements of actual use, OSNR monitoring is also explored, with a total precision equal to 95.2%. Furthermore, a thorough investigation of the impact of training data dimensions, on TL performance is conducted. The outcomes of the proposed method demonstrate that the suggested TL-based algorithms are more accurate and need much less training and testing time than non-TL approaches.

Time-frequency dual-dimension deep neural network (TF-DD-DNN)-based OSNR and XT monitoring in mode-division multiplexing systems

We propose a time-frequency dual-dimension deep neural network (TF-DD-DNN)-based optical performance monitoring (OPM) scheme to jointly monitor the optical signal-to-noise ratio (OSNR) and the mode-coupling induced crosstalk strength (XT) in mode-division multiplexing (MDM) systems, by revealing the diversity of the amplified spontaneous emission (ASE) and XT distortions on optical signals in the time- and frequency-domains. We numerically quantify the predictions of OSNR and XT for a coherent MDM transmission link, achieving the OSNR prediction accuracy of more than 0.99 and the XT prediction accuracy of up to 1 with the root mean square error (RMSE) of 0.11 dB at OSNR = 30 dB @0.1 nm for the case of quadrature-phase-shift keying (QPSK) signals. We evaluate the impacts of the chromatic dispersion (CD) of fibers and the low-pass filter bandwidth on the prediction procedures. The dependency of the modulation formats is also discussed, confirming the robust operation of the proposed scheme. Finally, the TF-DD-DNN scheme is experimental verified. High consistency is obtained between the simulation and experimental results.

Combined modulation format identification and optical signal-to-noise ratio monitoring with high accuracy and generalizability based on a proposed fused module few-shot learning algorithm in dynamic coherent optical transmissions

Modulation format identification (MFI) and optical signal-to-noise ratio (OSNR) monitoring are important portions of optical performance monitoring (OPM) for future dynamic optical networks. In this paper, we proposed a fusion module few-shot learning (FMFSL) algorithm as an improvement upon the ordinary few-shot learning algorithms for image recognition with the specialty in adopting a combination of a dilated convolutional group and an asymmetric convolutional group to advance the feature extraction. FMFSL algorithm is applied in MFI and OSNR monitoring in coherent optical communication systems with its performance investigated in both back-to-back and fiber transmission scenarios using small-scale constellation diagrams. The results show that FMFSL algorithm can achieve 100% accuracy in MFI and higher OSNR monitoring accuracy compared to the few-shot learning algorithms Deep Nearest Neighbor Neural Network (DN4) and Prototypical Nets (PN) with 2.14% and 4.28% for 64QAM and 3.38% and 8.06% for 128QAM, respectively, without much increase in time consumption. Furthermore, the trained FMFSL algorithm remains excellent in MFI and OSNR monitoring without retraining while employed in back-to-back transmission scenarios with smaller OSNR intervals and fiber transmission scenarios with different amounts of Kerr nonlinearity, demonstrating its high capabilities in generalization and robustness. FMFSL algorithm provides a potential solution for OPM in future dynamic optical networks as a novel machine learning tool.

Simultaneous OSNR and chromatic dispersion monitoring based on optical tones power ratio

A novel technique is proposed for simultaneous monitoring of optical signal-to-noise ratio (OSNR) and chromatic dispersion (CD). The scheme is based on the cross-phase modulation effect in a highly nonlinear fiber, which induces a change in the optical tones according to the impairments. The power ratio of the optical tones at two separate frequencies is utilized for monitoring. Improvements in the maximum measurable OSNR, CD and dynamic range can be obtained using only the clock tone for the non-return-to-zero, return-to-zero with a 50% duty cycle differential quadrature phase-shift keying, differential phase-shift keying and on-off keying signals at 80 Gb/s. For the return to zero differential quadrature phase shift keying (RZ-DQPSK) system, the OSNR of 4–40 dB can be monitored with a 36.73 dB dynamic range, and a chromatic dispersion of 0–117 ps/nm can be monitored with a 48.68 dB dynamic range. A correlation is built to describe the relationship of the power ratio with the OSNR and chromatic dispersion which has a root-mean-squared error of 1.41. The delay group dispersion insensitivity, the impact of the input signal power, and the filter selection are investigated. The design enables accurate monitoring, simple operation and may be used in a wide range.

Modulation Format Identification and OSNR Monitoring Based on Multi-Feature Fusion Network

In this paper, we propose a multi-feature fusion network (MFF-Net) for a modulation format identification (MFI) and optical signal-to-noise ratio (OSNR) monitoring scheme. The constellation map data used in this work comes from five modulation formats, namely 56 Gbit/s 4/8 phase shift keying (PSK) and 16/32/64 quadrature amplitude modulation (QAM). The constellation maps are input to one branch network of the MFF-Net, and then the constellation maps are processed by horizontal projection and used as input to another branch network as a way to fuse the two image features. The results show that the scheme achieves 100% MFI accuracy and 98.82% OSNR monitoring accuracy for the five modulation formats. In addition, the performance of MFF-Net and binarized convolutional neural network (B-CNN), visual geometry group network (VGG-Net), and traditional weighted multi-task learning (EW-MTL) are compared to present the superiority of the method. The effect of model structure on MFF-Net is also discussed. The robustness of the model is also evaluated for different transmission distances and bit rates.

Cascaded modulation format identification and optical signal-to-noise ratio estimation method for optical communication system

Modulation format identification (MFI) and optical performance monitoring are important for elastic optical networks. We propose a cascaded module-based MFI and optical signal-to-noise ratio (OSNR) estimation method. Polar coordinates were used as features to distinguish the modulation formats (MFs) and estimate the OSNR. Dilated convolution was adopted to increase the receptive field and reduce computational complexity. Four commonly used MFs were investigated: dual-polarization (DP) quadrature phase shift keying, 8QAM, 16QAM, and 32QAM. A 32G baud DP transmission system is established, and the results are discussed. The results show that MFI accuracy can reach 100% when the OSNR is less than the 7% forward error correction limit, and the accuracy of OSNR monitoring is close to 100%. Two factors, sample length and nonlinearity, were further investigated. The results show that the MFI and OSNR estimators can achieve perfect performance when the sample length is 3000. Simultaneously, the accuracy of the MFI remained at 100%, and the accuracy of the OSNR estimator decreased by 1% to 2% when the launch power changed from 0 to 5 dBm. Furthermore, two modules with the same structure, DC-CNN and CNN, were compared. The results show that the two models can achieve similar accuracies and that the DC-CNN has the least number of parameters. Finally, experimental verification was conducted to ensure the practical applicability of the proposed method.

Simultaneous modulation format identification and OSNR monitoring based on optoelectronic reservoir computing.

An approach for simultaneous modulation format identification (MFI) and optical signal-to-noise ratio (OSNR) monitoring in digital coherent optical communications is proposed based on optoelectronic reservoir computing (RC) and the signal's amplitude histograms (AHs) obtained after the adaptive post-equalization. The optoelectronic RC is implemented using a Mach-Zehnder modulator and optoelectronic delay feedback loop. We investigate the performance of the proposed model with the number of symbols, bins of AHs and the hyperparameters of optoelectronic RC. The results show that 100% MFI accuracy can be achieved simultaneously with accurate OSNR estimation for different modulation formats under study. The lowest achievable OSNR estimation mean absolute errors for the dual-polarization (DP)-quadrature phase-shift keying signal, the DP-16-ary quadrature amplitude modulation (16QAM) signal, and the DP-64QAM signal are 0.2 dB, 0.32 dB and 0.53 dB, respectively. The robustness of the proposed scheme is also evaluated when the optoelectronic RC is in presence of additive white Gaussian noises. Then, a proof of concept experiment is demonstrated to further verify our proposed method. The proposed approach offers a potential solution for next-generation intelligent optical performance monitoring in the physical layer.

Intelligent Optical Performance Monitoring Based on Intensity and Differential-Phase Features for Digital Coherent Receivers

An intelligent optical performance monitoring scheme for simultaneous modulation format identification (MFI) and optical signal-to-noise ratio (OSNR) monitoring is proposed and experimentally demonstrated in digital coherent receivers. The proposed scheme introduces convolutional neural network (CNN) to automatically extract and monitor modulation format and OSNR dependent features (the empirical distribution of intensity and differential-phase) which can be obtained after constant modulus algorithm (CMA) equalization. The experiment results show that 100% identification accuracies for all modulation formats (e.g. 28-GBaud PDM-QPSK/−8PSK/ −8QAM/−16QAM/−32QAM/−64QAM) are achieved at OSNR values are lower than the corresponding theoretical 20% FEC limit (BER = 2.4×10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−2</sup> ). Furthermore, under the chromatic dispersion (CD) variation from 0-ps/nm to 1200-ps/nm, the root-mean- square error (RMSE) and mean absolute error (MAE) of OSNR monitoring for all modulation formats are less than 0.0896-dB and 0.0657-dB, respectively. Subsequently, the influence of frequency offset and fiber nonlinearity effect on the intelligent optical performance monitoring scheme is also analyzed. We believe that our proposed multi-parameter monitoring scheme has the potential to be applied in the next-generation intelligent elastic optical networks.

Intelligent equally weighted multi-task learning for joint OSNR monitoring and modulation format identification

A novel scheme for combining modulation format identification (MFI) and optical performance monitoring (OPM) in elastic optical networks using convolutional-neural-network-based and equally weighted multi-task learning is proposed. In this scheme, shared features of optical signal-to-noise ratio (OSNR) and modulation formats (10 GBaud QPSK, QAM8, QAM16, QAM64) are extracted from constellation maps and used for multi-label classification. Simulation results demonstrate that the accuracies of MFI and OSNR can reach 100% stably and continuously. Moreover, we evaluated the robustness of this model in a system with fixed physical parameters and concluded that it is tolerant to chromatic dispersion and transmission distance. This technology predicts an image in ∼0.51 ms. Compared with the general multi-task learning (MTL) scheme, the proposed approach massively saves execution time (∼107 h) by removing weight selection. It is time-efficient and intelligent for OPM equipment that will perform multiple monitoring tasks in next-generation optical networks.

Meta-learning-enabled accurate OSNR monitoring of directly detected QAM signals with one-shot training

We experimentally demonstrate meta-learning-enabled accurate optical signal-to-noise ratio (OSNR) monitoring of directly detected 16QAM signals with extremely few training data. When one-shot training, where one amplitude histogram (AH) for each OSNR value includes only 2000 data samples, is implemented for a 16QAM signal within a variable OSNR range of 15-24 dB, the experimental root mean squared error (RMSE) of the retraining technique is 1.53 dB. For transfer learning from the 16QAM simulation to the experimentally generated AH, the RMSE can be reduced to 1.11 dB. In comparison with both the retraining and transfer learning techniques, the RMSE of meta-learning-enabled OSNR monitoring can be further reduced by 42.8% and 22.3%, respectively. In order to reach the optimal accuracy with an RMSE of 0.66 dB, the meta-learning technique requires only 15 AHs for each OSNR value to be monitored, while the retraining and the transfer learning techniques need 20 and 25 AHs, respectively.

Optical performance monitoring in transparent fiber-optic networks using neural networks and asynchronous amplitude histograms

Recently, technologies such as artificial neural networks (ANNs) and asynchronous amplitude histograms (AAHs) are widely employed in the optical performance monitoring (OPM). In this paper, the number of layers and the optimization of the neural networks are investigated to complete a series of monitoring tasks in optical communication systems with a large range of chromatic dispersion (CD) and optical signal-to-noise ratio (OSNR). It is shown that the monitoring accuracy has been significantly improved with such implementations. Numerical simulations, which have been conducted for six different modulation formats, demonstrate a good monitoring performance, with an average OSNR monitoring error of 0.1064 dB (99.49% monitoring accuracy) for the OSNR range of 10–40 dB and an average CD monitoring error of 3.3324 ps/nm (99.45% monitoring accuracy) for the CD range of 170–1870 ps/nm, respectively. Furthermore, the CD monitoring and the modulation format identification (MFI) have also been investigated for the system with a large range of dispersion. An average CD error of 16.6832 ps/nm (99.45% monitoring accuracy) and a 100% MFI accuracy have been achieved for all six modulation formats. This scheme is also verified experimentally with OSNR errors of 0.41 dB and CD errors of 15.21 ps/nm, respectively. The identification accuracy of the three modulation formats in the experimental system is also 100%.

Multi-Task Learning Convolutional Neural Network and Optical Spectrums Enabled Optical Performance Monitoring

We propose a simultaneous optical signal recognition (OSR) and optical signal-to-noise ratio (OSNR) monitoring method by using multi-task learning convolutional neural network (MTL-CNN) in conjunction with optical spectrums, which enables optical performance monitoring (OPM) in the optical transmission network. In order to achieve a trade-off between monitoring loss and time consumption of the MTL-CNN constructed for seven commonly used signals with a spectrum resolution of 10 pm, the number of feature map channels in four convolutional layers is delicately designed as 32, 32, 64, and 64 with task weights set to 0.4 and 0.2, corresponding to OSNR monitoring and OSR, respectively. Simulation results manifest that this method can realize the recognition of the received optical signals with an overall accuracy of 100% and also enable OSNR monitoring with the mean absolute error (MAE) of 0.262 dB. The proposed method shows strong robustness to various distortions and exhibits good performance in terms of time consumption. The effectiveness is further verified by proof-of-concept experiments in three signals. These illustrate that our method is a promising solution for multi-parameter monitoring with high accuracy and efficiency.

Joint multi-parameter optical performance monitoring scheme based on trajectory information for a Stokes vector direct detection system.

In this study, we propose and verify a joint multi-parameter optical performance monitoring (OPM) scheme based on trajectory information for the Stokes vector direct detection (SVDD) system, for the first time, to the best of our knowledge. Here, the proposed scheme first performs quantification of the trajectory to construct trajectory information, which not only presents diversity of the received symbols in spatial dimension, but also records the jump pattern among symbols in time dimension. Subsequently, eigenanalysis is introduced to extract critical features hidden in trajectory information and simultaneously achieve the purpose of dimensionality reduction. The effectiveness of the scheme is verified through 14/28 GBaud SVDD binary phase shift keying/quadrature phase shift keying/-8 quadrature amplitude modulation (QAM)/-16QAM/-32QAM/-64QAM simulation systems. Under the scenario of joint modulation format (MF) identification and optical signal to noise ratio (OSNR) monitoring, the identification rates of all six kinds of MFs achieve 100% within their corresponding reasonable OSNR ranges. Besides that, the average mean absolute error (MAE) of the monitored OSNRs are obtained as 0.03 dB, 0.22 dB, 0.36 dB, 0.41 dB, 0.46 dB, and 0.49 dB for those six kinds of MFs, respectively. Under the scenario of multi-parameter OPM, SVDD-8QAM/-16QAM/-32QAM signals are 100% successfully identified when residual chromatic dispersion (RCD) is located in the ranges of 0-200 ps/nm, 0-190 ps/nm, and 0-160 ps/nm, respectively. The average MAE of OSNR monitoring and RCD estimation for these three commonly used MFs are 1.08 dB and 3.23 ps/nm, respectively. Moreover, the study also demonstrates the robustness for baud rates and a relatively simpler calculation complexity about the proposed OPM scheme.

Optical performance monitoring using SOI-based spectral analysis.

A novel optical performance monitoring (OPM) method based on Fourier transform spectrum analysis (FTSA) is designed for optical signal-to-noise ratio (OSNR) monitoring, modulation format and baud rate recognition in the presence of fiber nonlinearities. The interference intensities, which reflect spectral features of signals, are obtained by exploiting the FTSA consisting of two-stage Mach-Zehnder interferometer (MZI) arrays. Then, the mapping between the OPM parameters and modulated interference intensity (MII) is characterized using neural networks without prior knowledge of the configuration of the communication network. Results show that optical performance parameters are monitored simultaneously. Meanwhile, the accuracy of modulation format and baud rate recognition is 94.8% and most (over 86%) OSNR monitoring errors are less than ±1 dB under complex transmission conditions in presence of frequency offset and delay jitter. Besides, the FTSA can be fabricated on a silicon on insulator (SOI) platform with a large fabrication tolerance, and it has broad working bandwidth to support the full optical communication band. Therefore, the proposed OPM method is capable of integration and miniaturization, which can be ubiquitously applied in network intermediate nodes to support the construction of smart optical networks.

Joint monitoring of OSNR and frequency offset for coherent optical fiber communication systems based on QPSK training sequence

We propose a joint monitoring scheme of optical signal-to-noise ratio (OSNR) and frequency offset (FO) based on the analysis of a quadrature phase-shift keying training sequence (QPSK-TS) spectrum for coherent optical fiber communication systems. In the proposed scheme, the FO estimation (FOE) consists of a coarse FOE stage and a fine FOE stage. Firstly, the coarse FOE stage is operated by searching for the peak of QPSK-TS spectrum. Then, after conducting coarse FO compensation (FOC) and down-sampling process to the QPSK-TS signal, the fine FOE stage can be realized by operating fast Fourier transform (FFT) to the 4th power of the signal. Compared with the traditional FOE scheme based on FFT with 4th-power operation, the FOE range of the proposed scheme can be extended to 3 times without losing the FOE resolution. After the FOC procedure with the proposed FOE scheme, OSNR can be monitored accurately by obtaining the whole in-band amplified spontaneous emission (ASE) noise from QPSK-TS spectrum. It is noteworthy that both the proposed FOE scheme and the OSNR monitoring scheme are insensitive to chromatic dispersion (CD) and intra-channel nonlinearity (NL). Simulation results for a 28 GBaud system with 2400 km transmission length show that when the OSNR range is 10∼28 dB and FO is 10.25 GHz, the mean OSNR estimated error is < 0.9 dB and the standard deviation is < 0.3 dB. Furthermore, the back-to-back (B2B) experiment results with 20 GBaud show that when the OSNR range is 10∼26 dB and FO is 7 GHz, the mean OSNR estimated error is < 0.8 dB and the standard deviation is also < 0.3 dB.

Simultaneous monitoring of residual chromatic dispersion and OSNR of NRZ-OOK signal based on asynchronous delay-tap sampling and image processing

We demonstrated a simultaneous residual chromatic dispersion (CD) and optical signal-to-noise ratio (OSNR) monitoring method for non-return-to-zero on-off keying signals by employing the delay-tap sampling and image processing techniques. Delay-tap sampling scatter plots reflect the signal pulse shape change and image processing can extract the contour of the scatter plots. Numerical simulation shows that the OSNR monitoring range is from 12 to 32 dB with <0.5 dB error and the CD monitoring range is from 0 to 1360 ps/nm. In the experiment, the OSNR monitoring range is from 14 to 30 dB and the CD monitoring range can go up to 1275 ps/nm.

Joint Modulation Format Identification and Optical Signal-to-Noise Ratio Monitoring Based on Ternary Neural Networks

Modulation format identification (MFI) and optical signal-to-noise ratio (OSNR) monitoring are essential for elastic optical networks. A method for joint modulation format identification and OSNR monitoring based on ternary neural networks was proposed. Further, a ternary neural network was established, and the constellation images after constant modulus algorithm (CMA) equalization were used as input features. Four commonly used modulation formats were distinguished, including dual-polarization (DP)-QPSK, 8QAM, 16QAM, and 64QAM. A 32G baud simulation system was constructed, and the results were analyzed. First, we investigated the influence of resolution and sample length on the identification accuracy. The values 32 and 7000 were selected for the two parameters, respectively, based on the balance between resource consumption and accuracy. Then, we compared the performance of the binary neural network (BNN), ternary neural network (TNN), and full-precision neural network (FNN). The results indicate that the memory consumption and extraction time of the TNN are similar to those of the BNN, and the accuracy is further improved. Moreover, the robustness of the system was analyzed. The results validate that the proposed method can tolerate fiber nonlinearity (from 500 km to 1500 km). Finally, an experiment was conducted to prove the practicality of the proposed method.

Accurate OSNR monitoring based on data-augmentation-assisted DNN with a small-scale dataset.

Deep neural networks (DNNs) have been successfully applied for accurate optical signal-to-noise ratio (OSNR) monitoring. However, the performance of OSNR monitoring substantially degrades when a mega dataset is inaccessible. Here, we demonstrate an accurate OSNR monitoring scheme based on a data-augmentation (DA)-assisted DNN with a small-scale dataset. When a 20 GBaud quadrature phase shift keying (QPSK) signal is transmitted over 400 to 2600 km standard single-mode fiber (SSMF) with an OSNR range from 8 to 14 dB, we experimentally evaluate the minimum dataset size to secure a mean absolute error (MAE) of OSNR monitoring less than 1 dB. The DA-assisted scheme only requires 50% of the raw data, in comparison with the traditional DNN scheme. Thus, the DA-assisted DNN scheme is promising for field-trial accurate OSNR monitoring, especially when the collection of mega datasets is inconvenient.

A Low-complexity OSNR monitoring scheme based on amplitude variance analysis

In this work, we have proposed and verified a low-complexity optical signal-to-noise ratio (OSNR) monitoring scheme based on amplitude variance analysis (AVA). The key feature for OSNR monitoring is extracted from the variance between the received symbols and the corresponding ideal amplitude circles. The simulation results have demonstrated that the obtained mean absolute errors (MAEs) of OSNR monitoring achieve 0.1176 dB, 0.2064 dB, 0.2179 dB, 0.2249 dB, and 0.2631 dB for 28 GBaud polarization-division-multiplexing (PDM) quadrature-phase-shift-keying (QPSK), PDM-8-quadrature-amplitude-modulation (QAM), PDM-16QAM, PDM-32QAM and PDM-64QAM, respectively. Meanwhile, the proposed AVA scheme presents good tolerances to chromatic dispersion (CD) and polarization-mode dispersion (PMD). Furthermore, the OSNR monitoring performances have been verified by a 28 GBaud PDM-QPSK/8QAM/16QAM/32QAM experimental transmission system. The obtained MAEs are 0.1293 dB, 0.3105 dB, 0.3214 dB, and 0.3823 dB, respectively. Compared with the moment estimation (MEB) and deep-neural-network (DNN)-based schemes, the accuracy of the proposed AVA scheme has been improved and its computation complexity is well reduced.

On-chip OSNR monitoring system based on power-balanced Mach-Zehnder interferometers

Optical signal-to-noise ratio (OSNR) monitor is a key unit in the modern optical networks, serving as an indicator of the transmission link quality. With the demand for cost efficiency, commercial systems require OSNR monitors with compact footprint, low cost, and less power consumption. In this paper, we demonstrate a fully integrated in-band optical signal-to-noise ratio (OSNR) monitor based on silicon-on-insulator (SOI) platform. Assembling the variable optical attenuators (VOAs) and photodetectors (PDs) together with power-balanced Mach-Zehnder interferometers (MZIs), the device has a compact footprint of 3 mm ×1 mm, and is fabricated through a commercial Multi-Project Wafer (MPW) run. 28 Gbaud NRZ-QPSK signal is used to characterize the device performance, and experimental results show errors within ±0.5 dB for OSNR range from 7.8 dB to 24.5 dB can be successfully achieved. Furthermore, the robustness of the tested device to signal power and chromatic dispersion is also proved. This monitor has potential to be deployed in the intermediation nodes of transmission links in a cost-effective way.