Related Experiment Video
Updated: Sep 20, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
652
Bi-LSTM-Augmented Deep Neural Network for Multi-Gbps VCSEL-Based Visible Light Communication Link
Seoyeon Oh1, Minseok Yu1, Seonghyeon Cho1
1Department of Information and Telecommunication Engineering, Incheon National University, Incheon 22012, Korea.
Sensors (Basel, Switzerland)
|June 10, 2022
Summary
Machine learning enhances vertical-cavity surface-emitting lasers (VCSELs) for faster optical wireless communications. A bidirectional long short-term memory (Bi-LSTM) model achieved 13.5 Gbps in visible light communication.
Area of Science:
- Optoelectronics
- Wireless Communications
- Machine Learning
Background:
- Vertical-cavity surface-emitting lasers (VCSELs) are promising for optical wireless communications.
- Off-the-shelf VCSELs have limited modulation bandwidth for multi-Gb/s data rates.
- Machine learning (ML) is emerging as a solution for communication challenges.
Purpose of the Study:
- To investigate the effectiveness of ML in enhancing VCSEL-based visible light communication (VLC).
- To demonstrate a high-speed VLC link using ML with commercial VCSELs.
Main Methods:
- Systematic analysis of ML techniques for VCSEL-based VLC.
- Implementation of a bidirectional long short-term memory (Bi-LSTM) ML scheme.
- Utilizing a commercial VCSEL device.
Main Results:
- Demonstrated a high-speed visible light communication (VLC) link.
- Achieved a data rate of 13.5 Gbps over a single channel.
- This represents the fastest reported single-channel result from a cost-effective, off-the-shelf VCSEL.
Conclusions:
- Machine learning, specifically Bi-LSTM, is highly effective in overcoming bandwidth limitations of VCSELs for VLC.
- The demonstrated 13.5 Gbps data rate signifies a significant advancement for cost-effective, high-speed optical wireless communication systems.
