Deep photonic reservoir computer for nonlinear equalization of 16-level quadrature amplitude modulation signals
Photonic reservoir computer (PRC) is a kind of real-time and adaptive recurrent neural network, where only weights in the readout layer require training. PRC is a promising tool to deal with the crucial issue of nonlinear equalization in optical fiber communications. Here, we theoretically present a...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIP Publishing LLC
2025-06-01
|
| Series: | APL Machine Learning |
| Online Access: | http://dx.doi.org/10.1063/5.0253655 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Photonic reservoir computer (PRC) is a kind of real-time and adaptive recurrent neural network, where only weights in the readout layer require training. PRC is a promising tool to deal with the crucial issue of nonlinear equalization in optical fiber communications. Here, we theoretically present a deep PRC for the nonlinear equalization of coherent signals with the format of 16-level quadrature amplitude modulation (16-QAM). The deep PRC consists of cascading injection-locked Fabry–Perot lasers with optical feedback. Both the in-phase component and the quadrature component of the 16-QAM signals are injected simultaneously into the deep PRC in parallel, based on the wavelength multiplexing of Fabry–Perot lasers. It is demonstrated that the deep PRC exhibits strong capabilities for the nonlinearity compensation of coherent signals. The Q factor is improved by more than 1 dB for 16-QAM signals with launch powers above 10 dBm, associated with a bit rate of 240 Gbps and a transmission distance of 50 km. |
|---|---|
| ISSN: | 2770-9019 |