Spiking Neural Network Linear Equalization: Experimental Demonstration of 2km 100Gb/s IM/DD PAM4 Optical Transmission

Georg Böcherer, Florian Strasser, Elias Arnold, Youxi Lin, Johannes Schemmel, Stefano Calabrò, Maxim Kuschnerov

2023 · 10 citations · 3 references

Concepts

Abstract

A linear feed-forward equalizer is implemented by a potentially low-power spiking neural network. For a 100Gb/s PAM-4 IM/DD optical 2km transmission, no performance penalty compared to a digital implementation is observed.

References

3