Wireless ultraviolet scattering channel estimation method based on deep learning.

Area of Science:

• Optical Communications
• Signal Processing
• Machine Learning

Background:

• Wireless ultraviolet (UV) communication faces challenges like inter-symbol interference (ISI) and signal attenuation in non-line-of-sight (NLOS) conditions due to strong scattering.
• Existing channel estimation methods struggle to effectively mitigate these issues in complex UV scattering environments.

Purpose of the Study:

• To develop an advanced channel estimation method for wireless UV communication systems operating in NLOS scattering environments.
• To improve the accuracy and reliability of UV communication by addressing ISI and transmission attenuation.

Main Methods:

• A novel deep learning model combining a one-dimensional convolutional neural network (1D-CNN) and a deep neural network (DNN) was designed for channel estimation.
• The differential evolution (DE) algorithm was employed to optimize the network training process, enhancing model performance.
• Computer simulations were used to evaluate the proposed method against traditional algorithms.

Main Results:

• The proposed deep learning scheme demonstrated superior mean square error (MSE) and bit error rate (BER) performance compared to conventional methods.
• The channel estimation method proved stable across diverse communication environments.
• The developed neural network model exhibited strong generalization capabilities.

Conclusions:

• The integrated 1D-CNN and DNN model, optimized with DE, offers a robust solution for wireless UV scattering channel estimation.
• This deep learning approach significantly enhances communication reliability and performance in challenging NLOS UV communication scenarios.
• The method's stability and generalization ability make it suitable for practical deployment in various wireless UV communication systems.