Nonlinear Digital Self-Interference Cancellation in In-Band Full-Duplex Systems with Complex-Valued Temporal...

Area of Science:

• Wireless communication systems
• Signal processing

Background:

• Full-duplex (FD) technology enables simultaneous transmission and reception, boosting spectral efficiency for integrated sensing and communication (ISAC).
• Non-ideal hardware in FD systems generates nonlinear self-interference (SI), hindering practical deployment.
• Traditional polynomial models for nonlinear SI have high computational complexity.

Purpose of the Study:

• To propose a novel method for nonlinear self-interference modeling and cancellation in FD systems.
• To address the limitations of conventional polynomial models in terms of complexity and performance.

Main Methods:

• A complex-valued temporal convolutional network (CV-TCN) with a wavelet activation function was developed for nonlinear SI modeling.
• The proposed CV-TCN was implemented as a canceller for nonlinear SI.
• Performance was evaluated through simulations, comparing it against polynomial-based and other cancellers.

Main Results:

• The CV-TCN canceller demonstrated superior nonlinear SI cancellation (SIC) performance compared to traditional methods.
• The proposed model achieved higher parameter efficiency.
• The CV-TCN suppressed residual SI power by approximately 7.95 dB after linear SIC, with residual SI remaining 2.23 dB above the noise floor.

Conclusions:

• The CV-TCN offers a computationally efficient and effective solution for nonlinear SI cancellation in FD systems.
• This advanced modeling approach significantly improves the practical feasibility of FD-based ISAC architectures.
• The proposed method enhances overall system performance by reducing self-interference below practical thresholds.