Constellation-based identification of linear and nonlinear OSNR using machine learning: a study of link-agnostic performance.

We demonstrate accurate estimation of generalized optical signal to noise ratio (GOSNR) for wavelength division multiplexed fiber communication systems using an experimentally trained multi-tasking convolutional neural network while simultaneously estimating linear and nonlinear noise contributions. Using dual-polarized 32-GBaud 16QAM DWDM links we extract learnable features from constellation density matrices and accurately estimate GOSNR while simultaneously estimating linear and nonlinear contributions. Estimation of the OSNRASE, OSNRNL and GOSNR are demonstrated with < 0.5 dB mean absolute error. We also assess the universality of our model within the regime of metro networks by cross-training with data from such links comprised of different fiber types. We demonstrate a path to a practical universal training method that includes additional link parameters. The methods do not require contiguous high-speed sampling, additional hardware nor transmission of special symbols or patterns and are readily implemented in deployed systems.

Optical performance monitoring using digital coherent receivers and convolutional neural networks.

We experimentally demonstrate accurate modulation format identification, optical signal to noise ratio (OSNR) estimation, and bit error ratio (BER) estimation of optical signals for wavelength division multiplexed optical communication systems using convolutional neural networks (CNN). We assess the benefits and challenges of extracting information at two distinct points within the demodulation process: immediately after timing recovery and immediately prior to symbol unmapping. For the former, we use 3D Stokes-space based signal representations. For the latter, we use conventional I-Q constellation images created using demodulated symbols. We demonstrate these methods on simulated and experimental dual-polarized waveforms for 32-GBaud QPSK, 8QAM, 16QAM, and 32QAM. Our results show that CNN extracts distinct and learnable features at both the early stage of demodulation where the information can be used to optimize subsequent stages and near the end of demodulation where the constellation images are readily available. Modulation format identification is demonstrated with >99.8% accuracy, OSNR estimation with <0.5 dB average discrepancy and BER estimation with percentage error of <25%.

Non-intrusive OSNR measurement of polarization-multiplexed signals with spectral shaping and subject to fiber non-linearity with minimum channel spacing of 37.5GHz.

A non-intrusive OSNR measurement technique relying on the detailed spectral comparison of an optical signal with its "noise-free" spectrum is described, including mathematical basis, validity conditions and algorithmic steps. The technique's performance is experimentally demonstrated with 100G PM-QPSK and 200G PM-16QAM signals subject to fiber non-linearity induced by 100G PM-QPSK and 10G NRZ-OOK neighbors. The OSNR measurement performance is also demonstrated when root-raised cosine spectral shaping is applied to the signals, with channel spacings of 50GHz and 37.5GHz. Experimental results for OSNR levels up to 30dB and launch powers up to 3dB above the optimum BER launch conditions are shown for different system and signal configurations.

Advanced signaling technologies for high-speed digital fiber-optic links.

We summarize the most recent research of the Georgia Tech Terabit Optical Networking Consortium and the state-of-the-art in fiber telecommunications. These results comprise high-capacity single-mode fiber systems with digital coherent receivers and shorter-reach multimode fiber links with vertical cavity surface emitting lasers. We strongly emphasize the capabilities that sophisticated digital signal processing and electronics add to these fiber-based data transport links.

Experimental study of cross-phase modulation reduction in hybrid systems with co-propagating 100G PM-QPSK and 10G OOK.

We experimentally investigate various methods for reducing cross-phase modulation in hybrid networks with mixed 100G and 10G traffic. The experimental results over standard single-mode and non-zero dispersion-shifted fiber types demonstrate the effectiveness of several different XPM reduction techniques as well as the interplay between them. Nonlinear transmission performance is quantified using the Nonlinear Threshold metric as a function of key system features, including DCM type, dispersion map, spectral guard bands, and carrier phase estimation window size. Fiber Bragg grating-based DCMs are shown to offer a distinct advantage over fiber-based DCMs under certain conditions, particularly in dispersion-managed systems with very strong XPM. The average walk-off per span is introduced as a simple yet effective metric to compare different methods of XPM mitigation.

Generalized weighted crosstalk for DWDM systems with cascaded wavelength-selective switches.

A generalized method for estimating transmission penalties from spectrally-shaped crosstalk in cascaded multi-port WSS networks is derived, including effects of ASE, signal filtering, and crosstalk filtering. The weighted crosstalk value is computed by multiplying the shaped interfering signal by the power spectral density of the primary signal. This value is used to predict OSNR penalties in networks with cascaded WSSs of arbitrary port count. Theoretical treatment is supported by extensive numerical simulations and experiments for a variety of network configurations. Examples are presented for 43 Gb/s DPSK and 120 Gb/s DP-QPSK in cascaded ROADM networks with three distinct WSS types.