Autoencoder assisted subcarrier optimization for nonlinear frequency division multiplexing.

Nonlinear frequency division multiplexing (NFDM) is a novel optical communication technique that can achieve nonlinear free transmission. However, current design of NFDM is analogous to orthogonal frequency division multiplexing (OFDM), where sinc function is utilized as subcarriers, which may not be optimal for nonlinear spectrums. In this paper, we propose an auto-encoder (AE) assisted subcarrier optimization scheme for dual-polarized (DP) NFDM systems. Numerical verifications show that our scheme can improve the Q-factor by 1.54 dB and 0.62 dB compared to sinc subcarrier and linear minimum mean square error (LMMSE) equalization, respectively, in a 960 km transmission scenario. We also analyze the characteristics of the optimized subcarriers and discuss how they enhance the performance. Furthermore, we demonstrate the robustness of the optimized subcarriers to different modulation formats, transmission distances and bandwidth. Our work provides a new idea in subcarrier design for NFDM.

Offset double sideband carrier assisted differential detection with field recovery at low carrier-to-signal power ratio.

Four system frameworks based on carrier assisted differential detection (CADD) receivers for offset double sideband (DSB) signal transmission, including offset DSB asymmetric CADD (offset DSB A-CADD), offset DSB symmetric CADD (offset DSB S-CADD), offset DSB parallel double delay asymmetric CADD (offset DSB PDD-A-CADD), and offset DSB parallel double delay symmetric CADD (offset DSB PDD-S-CADD) are proposed to reduce the requirement for carrier-to-signal power ratio (CSPR) and improve the spectral efficiency (SE) of the self-coherent detection. These frameworks accommodate signal-signal beat interference (SSBI) and efficiently solve the noise enhancement by placing a frequency gap as wide as the signal bandwidth in the middle of the left and right sideband signal. Massive theoretical derivation and simulation verification illustrated that compared with previous interleaved A-CADD, our system achieve field recovery under the condition of 0 dB CSPR with the improvement of SE by 5%, and the OSNR sensitivity is improved by 4.5 dB with 20% forward error correction (FEC) threshold. In addition, due to the devices' limited bandwidth (BW), the information-bearing signal is attenuated at the high-frequency region. And since SSBI has less influence on the signal in the high-frequency region, the frequency gap of the four offset DSB CADD schemes are compressed to utilize as much low-frequency resource as possible and improve the SE. Efficient compression of the frequency gap from 50% to 32.3% with 20% FEC threshold and 50% to 37.7% with 7% FEC threshold at 0 dB CSPR is achieved, and only a slight performance degradation is observed. At this time, the SE is improved by 22.7% and 17.3% with different FEC thresholds, respectively, compared with the 5% frequency gap interleaved A-CADD.

Joint nonlinear optical signal-to-noise ratio estimation and modulation format identification based on constellation-points trajectory information and multitask 1DCNN for WDM systems.

We propose a joint monitoring scheme of nonlinear optical signal-to-noise ratio (O S N R N L ) estimation and modulation format identification (MFI) in wavelength division multiplexing (WDM) systems. Based on the abundant information of both nonlinear noise (NLN) and modulation format (MF) in received signals, this scheme first counts the trajectory information of all adjacent constellation points, and then quantifies them into the adjacent matrix. Subsequently, the eigenvectors corresponding to the largest eigenvalues are extracted via eigen-decomposition of the adjacent matrix, which characterize the information of NLN and MF effectively. Finally, the eigenvectors are fed into multitask one-dimensional convolutional neural network to perform O S N R N L estimation and MFI simultaneously. To verify the effectiveness of the scheme, five-channel 28 GBaud polarization division multiplexing (PDM) -16/32/64 quadrature amplitude modulation (QAM) WDM simulation systems are built by VPI. The simulation results demonstrate that, for PDM-16/32/64QAM signals, the mean absolute errors of O S N R N L estimation are 0.18, 0.17, and 0.20 dB, respectively. At the same time, the identification accuracy rates of these three MFs have achieved 100% within the ranges of estimated O S N R N L . Furthermore, a three-channel 28 GBaud WDM experimental system is constructed to further investigate the effectiveness of trajectory information for O S N R N L estimation. The experimental results show that the O S N R N L estimation errors of PDM-16QAM are less than 0.5 dB. In addition, our analysis of complexity from two aspects of trajectory information extraction and neural network model shows that the overall complexity scale of this scheme is O(K i,3 M C i,3 C o,3).

Blind modulation format identification based on improved PSO clustering in a 2D Stokes plane.

Blind modulation format identification (MFI) is indispensable for correct signal demodulation and optical performance monitoring in future elastic optical networks (EON). Existing MFI schemes based on a clustering algorithm in Stokes space have gained good performance, while only limited types of modulation formats could be correctly identified, and the complexities are relatively high. In this work, we have proposed an MFI scheme with a low computational complexity, which combines an improved particle swarm optimization (I-PSO) clustering algorithm with a 2D Stokes plane. The main idea of I-PSO is to add a new field of view on each particle and limit each particle to only communicate with its neighbor particles, so as to realize the correct judgment of the number of multiple clusters (local extrema) on the density images of the s2-s3 plane. The effectiveness has been verified by 28 GBaud polarization division multiplexing (PDM)-BPSK/PDM-QPSK/PDM-8QAM/PDM-16QAM/PDM-32QAM/PDM-64QAM simulation EON systems and 28 GBaud PDM-QPSK/PDM-8QAM/PDM-16QAM/PDM-32QAM proof-of-concept transmission experiments. The results show that, using this MFI scheme, the minimum optical signal-to-noise ratio (OSNR) values to achieve 100% MFI success rate are all equal to or lower than those of the corresponding 7% forward error correction (FEC) thresholds. At the same time, the MFI scheme also obtains good tolerance to residual chromatic dispersion and differential group delay. Besides that, the proposed scheme achieves 100% MFI success rate within a maximum launch power range of -2∼+6 dBm. More importantly, its computational complexity can be denoted as O(N).