Transformer model improves digital interference cancellation in full-duplex radios

Digital Self-Interference Cancellation for In-Band Full-Duplex Systems Based on Transformer Neural Operators

Information TheoryMachine Learning

Summary

In full-duplex radios that send and receive signals at the same time, a device's own transmission can create interference that drowns out incoming signals. This interference is hard to remove, especially when the desired incoming signal is strong. The authors propose a new approach using a special kind of neural network called a Transformer to learn patterns in how this interference changes over time. Unlike traditional methods, this model is trained without the desired signals present and can then remove interference more reliably when those signals are there. Tests show this method reduces interference better in various channel conditions than older techniques.

What this means in practice

  • For wireless communication engineers: Enhance full-duplex radio systems by more effectively removing self-generated interference with a learned Transformer model, improving signal quality.
  • For telecommunication system integrators: Integrate Transformer-based digital cancellers to maintain stable interference cancellation across varying signal conditions and channels in network hardware.

Authors

Zhijie Su, Shiquan Zhang, Qiaolin He

Abstract

Digital-domain self-interference cancellation is essential in in-band full-duplex radios after passive suppression and analog cancellation. It is particularly challenging when the desired signal is strong relative to the residual self-interference. In this regime, conventional iterative digital cancellers update their coefficients from desired-signal-contaminated received samples and may suffer substantial performance degradation. This paper reformulates digital-domain self-interference cancellation as an operator-learning problem. The proposed framework learns the temporal evolution of the residual self-interference channel from historical observations and uses the learned operator together with the known local transmit signal to reconstruct the current residual self-interference. A Transformer neural operator is used to realize this framework. The resulting method is trained in a desired-signal-free setting and then applied in the presence of a desired signal. A continuity-based analysis characterizes the sensitivity of the learned operator to the resulting input perturbation. Numerical experiments in static and time-varying channels, including time-varying SIR conditions, demonstrate stable cancellation performance over a wide range of interference levels. At 30 dB SIR, the proposed method improves the cancellation performance by 5--15 dB over the considered conventional methods.
View PDFOpen arXiv