A Two-Stage Time-Domain Equalization Method for Mitigating Nonlinear Distortion in Single-Carrier THz Communication...

Area of Science:

• Wireless Communication
• Signal Processing
• Machine Learning

Background:

• Terahertz (THz) communication is crucial for high-speed data transmission due to its large bandwidth.
• Nonlinear distortions and random interference degrade THz communication performance.
• Effective equalization techniques are needed to enhance reliability.

Purpose of the Study:

• To propose a two-stage deep learning-based time-domain equalization method for single-carrier THz systems.
• To mitigate nonlinear distortions and improve communication reliability.
• To validate the method's performance and generalization capability.

Main Methods:

• A two-stage deep learning (DL) framework for time-domain equalization.
• A progressive learning strategy capturing global then local channel characteristics.
• Experimental validation at 230 GHz (2.1 m) and 310 GHz (1.5 m).

Main Results:

• Significant reduction in bit error rate (BER) was achieved.
• Approximately 92.15% BER reduction at 230 GHz and 83.33% at 310 GHz.
• Stable performance demonstrated across varying experimental conditions.

Conclusions:

• The proposed two-stage DL equalizer effectively mitigates nonlinear distortions in THz communication.
• The method offers a promising solution for enhancing THz communication reliability.
• This approach supports future research and development in THz wireless systems.