Samples vs. Symbols-based Feedforward Neural Network equalization for Short Reach Transmission

Yevhenii Osadchuk,Roman Dischler,Darko Zibar,Stenio M. Ranzini,Vahid Aref,Francesco Da Ros

2022 IEEE Photonics Conference (IPC)（2022）

引用 0|浏览5

暂无评分

摘要

We numerically and experimentally compare the performance of samples (Sa-FNN) versus symbols-based (Sy-FNN) feedforward neural network (FNN) equalizers to mitigate chromatic dispersion in intensity-modulated directly detected 32GBd OOK transmission with optical preprocessing. We demonstrate that Sy-FNN outperforms the Sa-FNN, increasing the reach by nearly 50%.

查看译文

关键词

Machine learning,Feedforward neural networks,Optical communications,IM/DD

AI 理解论文

溯源树

样例

生成溯源树，研究论文发展脉络

Chat Paper

正在生成论文摘要