Abstract

In soliton communication systems, the continuous nonlinear spectrum, ideally zero, is conventionally ignored at the receiver. In this paper, we exploit correlation between the received continuous spectrum and perturbations of the discrete soliton eigenvalue. We propose four estimation schemes, classified into two categories, one based on the nonlinear Fourier transform (NFT) and the other based on minimum Euclidean distance. Both categories comprise two schemes, one that exploits the received continuous spectral function to achieve improved estimation and one that does not. Numerical simulations demonstrate that significant reduction in estimation error can be achieved when the continuous spectrum is exploited, translating into improved information transmission rates of up to ${\text{46}}\%$ compared to the reference NFT-based scheme.

© 2019 IEEE

Neural network-aided receivers for soliton communication impaired by solitonic interaction

Yu Chen, Mohammadamin Baniasadi, and Majid Safari
Opt. Express 31(26) 43289-43306 (2023)

Improvement for a full-spectrum modulated nonlinear frequency division multiplexing transmission system

Jiacheng Wei, Lixia Xi, Xulun Zhang, Jiayun Deng, Ruofan Zhang, Shucheng Du, Wenbo Zhang, and Xiaoguang Zhang
Opt. Express 30(17) 31195-31208 (2022)

Signaling on the continuous spectrum of nonlinear optical fiber

Iman Tavakkolnia and Majid Safari
Opt. Express 25(16) 18685-18702 (2017)

Robust neural network receiver for multiple-eigenvalue modulated nonlinear frequency division multiplexing system

Yue Wu, Lixia Xi, Xulun Zhang, Zibo Zheng, Jiacheng Wei, Shucheng Du, Wenbo Zhang, and Xiaoguang Zhang
Opt. Express 28(12) 18304-18316 (2020)

Noise equalization scheme based on complex-valued ANN for multiple-eigenvalue modulated nonlinear frequency division multiplexing systems

Xiuhua Lv, Chenglin Bai, Qi Qi, Hengying Xu, Xueyuan Luo, Xinyu Chi, Lishan Yang, and LiXia Xi
Appl. Opt. 61(36) 10755-10765 (2022)