AI-Driven Real-Time Phase Optimization for Energy Harvesting-Enabled Dual-IRS Cooperative NOMA Under...

Area of Science:

• Wireless Communication Networks
• Signal Processing
• Machine Learning Applications

Background:

• Next-generation wireless networks require enhanced spectral efficiency and energy efficiency.
• Non-line-of-sight (NLoS) communication presents significant challenges in wireless systems.
• Traditional optimization methods for intelligent reflecting surfaces (IRSs) can be complex and computationally intensive.

Purpose of the Study:

• To propose a novel wireless network architecture integrating double IRSs, energy harvesting (EH), and cooperative NOMA (C-NOMA).
• To develop a low-complexity, learning-based approach for optimizing IRS phase shifts in dynamic NLoS environments.
• To evaluate the performance improvements in spectral efficiency and service reliability for green communication systems.

Main Methods:

• A wireless network architecture combining double IRSs, EH, and C-NOMA was designed.
• A lightweight machine learning model was employed to predict optimal IRS phase shift configurations.
• The learning-based model was trained on channel behavior and system observations for rapid adaptation.
• Numerical analysis was conducted to validate the proposed system's performance.

Main Results:

• The proposed architecture demonstrated significant improvements in spectral efficiency compared to conventional systems.
• Service reliability was considerably enhanced through the integration of EH and relay-based communication.
• The learning-based IRS phase shift optimization proved effective and low-complexity.
• The system facilitates the development of green communication systems.

Conclusions:

• The integrated architecture effectively leverages IRSs, EH, and C-NOMA for superior NLoS communication performance.
• The machine learning approach provides an efficient solution for IRS phase shift optimization in dynamic wireless environments.
• The proposed system contributes to achieving energy-efficient and reliable next-generation wireless networks.