Recurrent Neural Network Equalizer to Extend Input Power Dynamic Range of SOA in 100Gb/s/λ PON

Stephen Murphy,Paul D. Townsend,Cleitus Antony

Conference on Lasers and Electro-Optics（2022）

引用 1|浏览11

暂无评分

摘要

We propose a novel equalization scheme for 100Gb/s/λ PAM4 PON based on Gated Recurrent Neural Network to increase SOA preamplifier input power dynamic range tolerance to 30 dB below hard-decision FEC BER limit of 3.8×10−3.

查看译文

关键词

Recurrent Neural Network equalizer,novel equalization scheme,PAM4 PON,Gated Recurrent Neural Network,SOA preamplifier input power dynamic range tolerance,noise figure 30.0 dB

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