DOI | Resolve DOI: https://doi.org/10.1109/IPC53466.2022.9975752 |
---|
Author | Search for: Masnad, Md Mahadi; Search for: Gostimirovic, Dusan; Search for: Grinberg, Yuri1; Search for: Xu, Dan-Xia2; Search for: Liboiron-Ladouceur, Odile |
---|
Affiliation | - National Research Council of Canada. Digital Technologies
- National Research Council of Canada. Advanced Electronics and Photonics
|
---|
Format | Text, Article |
---|
Conference | 2022 IEEE Photonics Conference (IPC), November 13-17, 2022, Vancouver, BC, Canada |
---|
Subject | silicon photonics; topological optimization; machine learning; convolutional neural networks; nanofabrication; optical losses; fabrication; degradation; deep learning; neural networks; crosstalk; insertion loss |
---|
Abstract | We present a machine learning model to correct for fabrication variations in a topologically optimized mode demultiplexer and minimize its fabrication-induced performance degradation. The corrected design promises ~51% reduction in the insertion loss and an average reduction of crosstalk by 6 dB in simulation. |
---|
Publication date | 2022-11-13 |
---|
Publisher | IEEE |
---|
Licence | |
---|
In | |
---|
Language | English |
---|
Peer reviewed | Yes |
---|
Export citation | Export as RIS |
---|
Report a correction | Report a correction (opens in a new tab) |
---|
Record identifier | 16aa733b-d849-4863-9965-6a452130b436 |
---|
Record created | 2023-05-15 |
---|
Record modified | 2023-05-16 |
---|