Presentation | 2023-02-17 Designing Operation Parameters Using Deep Unfolding for MIMO Adaptive Equalization in Optical Coherent Transmission Systems Eidai Sunamoto, Koji Igarashi, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In the coherent optical transmission systems, adaptive control is required to suppress and compensate for laser phase noise and polarization rotation which are dynamically fluctuated. The large difference between their fluctuation speeds makes the MIMO adaptive equalization unstable. In particular, it is remarkable for quadrature amplitude modulation (QAM) optical signals. Although the improved techniques for MIMO adaptive equalization applicable to QAM signals have been proposed, their performance comparison is a critical issue because huge time and cost are required for exhaustive optimization fo design parameters when the linewidth of laser phase noise and signal-to-noise ratio are varied. In this work, we propose the optimization method using deep learning, called deep unfolding, to apply to the exhaustive parameter optimization in MIMO adaptive equalization. Using open libraries for deep learning, the cost can be remarkably reduced because full-scratch implement is not needed. Using our proposed method, we optimize the design parameters and evaluate the phase tracking performance for adaptive equalization in the conventional MIMO and the MIMO with separated phase compensation. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | optical coherent reception / digital signal processing / deep learning |
Paper # | OCS2022-81,OPE2022-110 |
Date of Issue | 2023-02-09 (OCS, OPE) |
Conference Information | |
Committee | OCS / OFT / OPE |
---|---|
Conference Date | 2023/2/16(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | kukuru-itomancity |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Takeshi Hoshida(Fujitsu) / Makoto Yamada(Osaka Metropolitan Univ.) / Toshikazu Hashimoto(NTT) |
Vice Chair | / / Taro Arakawa(Yokohama National Univ.) |
Secretary | (NTT) / (Sumitomo Electric Industries) / Taro Arakawa(Ryutsu Keizai Univ.) |
Assistant | / Yukihiro Goto(NTT) / Naoyuki Ozawa(Fujikura) / Yuhei Ishizaka(Kanto Gakuin Univ.) / Takeshi Umeki(NTT) |
Paper Information | |
Registration To | Technical Committee on Optical Communication Systems / Technical Committee on Optical Fiber Technology / Technical Committee on OptoElectronics |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Designing Operation Parameters Using Deep Unfolding for MIMO Adaptive Equalization in Optical Coherent Transmission Systems |
Sub Title (in English) | |
Keyword(1) | optical coherent reception |
Keyword(2) | digital signal processing |
Keyword(3) | deep learning |
1st Author's Name | Eidai Sunamoto |
1st Author's Affiliation | Osaka University(Osaka Univ.) |
2nd Author's Name | Koji Igarashi |
2nd Author's Affiliation | Osaka University(Osaka Univ.) |
Date | 2023-02-17 |
Paper # | OCS2022-81,OPE2022-110 |
Volume (vol) | vol.122 |
Number (no) | OCS-379,OPE-381 |
Page | pp.pp.56-61(OCS), pp.81-86(OPE), |
#Pages | 6 |
Date of Issue | 2023-02-09 (OCS, OPE) |