ASIC Implementation of Time-Domain Digital Backpropagation with\n Deep-Learned Chromatic Dispersion Filters

2018/06/19 by Christoffer Fougstedt, Fougstedt, Christoffer, Christian Häger +8
Engineering · #Optical Network Technologies #Photonic and Optical Devices #Advanced Photonic Communication Systems

paper · pdf · doi:10.48550/arxiv.1806.07223

Abstract

We consider time-domain digital backpropagation with chromatic dispersion\nfilters jointly optimized and quantized using machine-learning techniques.\nCompared to the baseline implementations, we show improved BER performance and\n>40% power dissipation reductions in 28-nm CMOS.\n

Related

• Wideband Time-Domain Digital Backpropagation via Subband Processing and\n Deep Learning
• Analog-to-digital conversion revolutionized by deep learning
• FPGA Implementation of Complex Value-based Clustering Filter for Chromatic Dispersion Compensation in Coherent Metro Links with Ultra-low Power Consumption
• FPGA Implementation of Multi-Layer Machine Learning Equalizer with On-Chip Training
• Recent Advances on Machine Learning-aided DSP for Short-reach and Long-haul Optical Communications