We experimentally demonstrated a novel nonlinearity mitigation scheme based on digital signal processing using a three-layer neural network (NN). 40-Gbit/s optical 16QAM signal distorted by SPM was compensated, improving EVM values by about 15%. We also performed numerical simulation of the proposed scheme, and confirmed that the experiment agrees with the results of the simulation. We performed 100 times of learning processes to find weight and bias of each neuron. However, we did not observe any serious local minimum. We also investigated the effect of the number of neurons of the NN on the compensation performance. Keywords: digital signal processing, nonlinear distortion, SPM, neural network Classification: Fiber-Optic Transmission for Communications , vol. 46, no. 16, pp. 1140-1141, Aug. 2010. DOI:10.1049/el.2010.1444 L. Liu, L. Li, Y. Huang, K. Cui, Q. Xiong, and F. N. Hauske, "Intrachannel nonlinearity compensation by inverse volterra series transfer function," J.
References
We experimentally demonstrate a novel nonlinearity-mitigation scheme based on a complex-valued neural network (CVNN) which is constructed by artificial neurons with complex-valued input and output. The in-phase (I) and quadrature (Q) components of optical signal are operated as complex values in the CVNN. A 40-Gbit/s optical 16QAM signal distorted by SPM was successfully compensated, improving error vector magnitude (EVM) by about 15%. The learning speed of the nonlinear equalizer was improved by using the CVNN, compared with conventional real-valued neural network (RVNN). Furthermore, the study show that CVNN has the potential to improve the computational complexity of RVNN.
The authors investigated the problem of overestimation with the Volterra series transfer function (VSTF) and an artificial neural network (ANN), which are used for non-linear equalisers in optical communication systems. The results revealed that the risk of predicting a pseudo-random binary sequence (PRBS) pattern, which causes overestimation of the equaliser performance, occurs not only with an ANN but also with the VSTF. When using PRBS9, PRBS11 and PRBS15, the number of taps of a feedforward tapped delay line, which is required in the VSTF to predict the PRBS pattern, was the same as that with the ANN. When the second-order Volterra kernels were omitted, a larger number of taps was required in the VSTF to observe the overestimation. a EVM versus the number of taps in ANN-based non-linear equaliser b EVM versus the number of taps in VSTF-based non-linear equaliser using first-, second-and third-order Volterra kernels c EVM versus the number of taps in VSTF with first-, second-and third-order Volterra kernels and VSTF with firstand third-order Volterra kernels (the VSTFs were trained on PRBS9) Conclusion: We investigated the problem of overestimation in the ANN-and VSTF-based non-linear equalisers in PRBS-based signal quality evaluation. The results revealed that a VSTF can predict PRBS patterns, and the overestimation problem occurs not only with the ANN but also with the VSTF. In our investigation using PRBS9, PRBS11 and PRBS15, the number of taps that the VSTF required to predict the PRBS pattern was the same as that with the ANN.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.