论文信息 - Towards Low Computational Complexity for Neural Network-Based Equalization in PAM4 Short-Reach Direct Detection Systems by Multi-Symbol Prediction

Towards Low Computational Complexity for Neural Network-Based Equalization in PAM4 Short-Reach Direct Detection Systems by Multi-Symbol Prediction

We propose NN-based multi-symbol equalization for a 50-Gb/s PAM4 direct-detection link which significantly reduces computational complexity without sacrificing BER performance of conventional single-output NNs. Complexity reduction of 42.7% is achieved when 8 outputs are selected.

Zhaopeng Xu | Chuanbowen Sun | William Shieh | Jonathan H. Manton | Shuangyu Dong

[1] Zhaopeng Xu,et al. Cascade recurrent neural network-assisted nonlinear equalization for a 100 Gb/s PAM4 short-reach direct detection system. , 2020, Optics letters.

[2] Chuanbowen Sun,et al. Computational complexity comparison of feedforward/radial basis function/recurrent neural network-based equalizer for a 50-Gb/s PAM4 direct-detection optical link. , 2019, Optics express.

[3] Jing Zhang,et al. Loss weight adaptive multi-task learning based optical performance monitor for multiple parameters estimation. , 2019, Optics express.

[4] Songnian Fu,et al. Transfer learning simplified multi-task deep neural network for PDM-64QAM optical performance monitoring. , 2020, Optics express.

[5] Ali Ghiasi,et al. Large data centers interconnect bottlenecks. , 2015, Optics express.

[6] Tao Gui,et al. Experimental study of PAM-4, CAP-16, and DMT for 100 Gb/s short reach optical transmission systems. , 2015, Optics express.

[7] Jason Weston,et al. A unified architecture for natural language processing: deep neural networks with multitask learning , 2008, ICML '08.