KRLS post-distorter with adaptive kernel width for visible light communications.

Area of Science:

• Wireless Communication
• Optical Engineering
• Signal Processing

Background:

• Visible Light Communication (VLC) is a key technology for high-speed data transmission, supplementing 5G wireless networks.
• Nonlinear characteristics of Light Emitting Diodes (LEDs) introduce signal distortion, degrading VLC performance.
• Existing linearization techniques may face limitations in adaptive learning and memory efficiency.

Purpose of the Study:

• To propose and evaluate a novel post-distortion scheme for mitigating LED nonlinearities in VLC systems.
• To enhance the adaptive learning capability and memory efficiency of linearization algorithms.
• To improve the overall communication quality and data transmission rates in VLC links.

Main Methods:

• A reproducing kernel Hilbert space (RKHS) based post-distortion scheme utilizing Kernel Recursive Least Squares (KRLS) with an adaptive kernel width.
• Integration of the Gauss-Newton (GN) algorithm to dynamically update the kernel width during the recursive KRLS process.
• Incorporation of the Enhanced Novelty Criterion (ENC) for adaptive sparse dictionary learning, optimizing linearization under memory constraints.

Main Results:

• The proposed KRLS-GN-ENC scheme demonstrates significant performance improvement compared to the KRLS-ENC method.
• KRLS-GN-ENC achieves superior nonlinear compensation compared to classical polynomial filtering techniques.
• The scheme exhibits faster convergence speeds in its linearization process.

Conclusions:

• The KRLS-GN-ENC scheme effectively mitigates nonlinear impairments in VLC systems.
• Adaptive kernel width and sparse dictionary learning enhance linearization performance and efficiency.
• This method offers a promising solution for high-quality, high-speed VLC.