An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communications
Intensity modulation/direct detection (IM/DD) remains to be the preferred optical transmission scheme for short-range applications for its simplicity of application, inexpensiveness, and small footprint. However, the impairments of low-cost device and fiber chromatic dispersion lead to the limitatio...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
AIMS Press
2024-09-01
|
| Series: | AIMS Electronics and Electrical Engineering |
| Subjects: | |
| Online Access: | https://www.aimspress.com/article/doi/10.3934/electreng.2024019 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832590267392196608 |
|---|---|
| author | Vuong Quang Phuoc Nguyen Van Dien Ho Duc Tam Linh Nguyen Van Tuan Nguyen Van Hieu Le Thai Son Nguyen Tan Hung |
| author_facet | Vuong Quang Phuoc Nguyen Van Dien Ho Duc Tam Linh Nguyen Van Tuan Nguyen Van Hieu Le Thai Son Nguyen Tan Hung |
| author_sort | Vuong Quang Phuoc |
| collection | DOAJ |
| description | Intensity modulation/direct detection (IM/DD) remains to be the preferred optical transmission scheme for short-range applications for its simplicity of application, inexpensiveness, and small footprint. However, the impairments of low-cost device and fiber chromatic dispersion lead to the limitation of system performance when the data rate rises to 100 Gbps or higher. In this paper, we demonstrated that an equalizer using neural networks can effectively improve the transmission performance of high-speed IM/DD systems. An optimization of a long short-term memory (LSTM) structure in terms of network depth and distribution of neurons in hidden layers leads to an enhancement of the overall performance of the 50 Gbaud PAM4 communications. Furthermore, the results for a system using a LSTM-based equalizer give the better outcome than the traditional feed-forward equalizer (FFE) or artificial neural network (ANN)-based equalizer. |
| format | Article |
| id | doaj-art-875bc28fa3d9493ba276495e5919d8b0 |
| institution | Kabale University |
| issn | 2578-1588 |
| language | English |
| publishDate | 2024-09-01 |
| publisher | AIMS Press |
| record_format | Article |
| series | AIMS Electronics and Electrical Engineering |
| spelling | doaj-art-875bc28fa3d9493ba276495e5919d8b02025-01-24T01:11:32ZengAIMS PressAIMS Electronics and Electrical Engineering2578-15882024-09-018439440910.3934/electreng.2024019An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communicationsVuong Quang Phuoc0Nguyen Van Dien1Ho Duc Tam Linh2Nguyen Van Tuan3Nguyen Van Hieu4Le Thai Son5Nguyen Tan Hung6The University of Danang - Advanced Institute of Science and Technology, Danang 50000, VietnamFPT University, Danang 50000, VietnamUniversity of Sciences, Hue University, Hue 49000, VietnamThe University of Danang - University of Science and Technology, Danang 50000, VietnamThe University of Danang - University of Science and Technology, Danang 50000, VietnamNubis Communications, New Providence NJ 07974, USThe University of Danang - Advanced Institute of Science and Technology, Danang 50000, VietnamIntensity modulation/direct detection (IM/DD) remains to be the preferred optical transmission scheme for short-range applications for its simplicity of application, inexpensiveness, and small footprint. However, the impairments of low-cost device and fiber chromatic dispersion lead to the limitation of system performance when the data rate rises to 100 Gbps or higher. In this paper, we demonstrated that an equalizer using neural networks can effectively improve the transmission performance of high-speed IM/DD systems. An optimization of a long short-term memory (LSTM) structure in terms of network depth and distribution of neurons in hidden layers leads to an enhancement of the overall performance of the 50 Gbaud PAM4 communications. Furthermore, the results for a system using a LSTM-based equalizer give the better outcome than the traditional feed-forward equalizer (FFE) or artificial neural network (ANN)-based equalizer.https://www.aimspress.com/article/doi/10.3934/electreng.2024019fiber-optic communicationsintensity modulationdirect detectionsignal equalizationlong short-term memoryartificial neural network |
| spellingShingle | Vuong Quang Phuoc Nguyen Van Dien Ho Duc Tam Linh Nguyen Van Tuan Nguyen Van Hieu Le Thai Son Nguyen Tan Hung An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communications AIMS Electronics and Electrical Engineering fiber-optic communications intensity modulation direct detection signal equalization long short-term memory artificial neural network |
| title | An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communications |
| title_full | An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communications |
| title_fullStr | An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communications |
| title_full_unstemmed | An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communications |
| title_short | An optimized LSTM-based equalizer for 100 Gigabit/s-class short-range fiber-optic communications |
| title_sort | optimized lstm based equalizer for 100 gigabit s class short range fiber optic communications |
| topic | fiber-optic communications intensity modulation direct detection signal equalization long short-term memory artificial neural network |
| url | https://www.aimspress.com/article/doi/10.3934/electreng.2024019 |
| work_keys_str_mv | AT vuongquangphuoc anoptimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyenvandien anoptimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT hoductamlinh anoptimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyenvantuan anoptimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyenvanhieu anoptimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT lethaison anoptimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyentanhung anoptimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT vuongquangphuoc optimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyenvandien optimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT hoductamlinh optimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyenvantuan optimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyenvanhieu optimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT lethaison optimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications AT nguyentanhung optimizedlstmbasedequalizerfor100gigabitsclassshortrangefiberopticcommunications |