A Novel Machine Learning-based Equalizer for a Downstream 100G PAM-4 PON

Optical Fiber Communication Conference, pp. W1H.1

Abstract

A frequency-calibrated SCINet (FC-SCINet) equalizer is proposed for downstream 100G PON with 28.7 dB path loss. At 5 km, FC-SCINet improves the BER by 88.87% compared to FFE and a 3-layer DNN with 10.57% lower complexity.

Authors 6

1. Chen Shao corresponding

   Karlsruhe Institute of Technology

   Affiliation as printed

   Karlsruhe Institute of Technology (KIT), Kaiserstr. 12, 76131 Karlsruhe, Germany

2. Elias Giacoumidis corresponding

   VPIphotonics (Germany)

   Affiliation as printed

   VPIphotonics GmbH, Hallerstr. 6, 10587 Berlin, Germany

3. Jialei Li corresponding Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University, Templergraben 57, 52062 Aachen

4. Michael Färber corresponding

   Karlsruhe Institute of Technology

   Affiliation as printed

   Karlsruhe Institute of Technology (KIT), Kaiserstr. 12, 76131 Karlsruhe, Germany

5. Tobias Käfer corresponding

   Karlsruhe Institute of Technology

   Affiliation as printed

   Karlsruhe Institute of Technology (KIT), Kaiserstr. 12, 76131 Karlsruhe, Germany

6. André Richter corresponding

   VPIphotonics (Germany)

   Affiliation as printed

   VPIphotonics GmbH, Hallerstr. 6, 10587 Berlin, Germany