A digital self-interference cancellation method for practical full-duplex radio

TLDR

Full‑duplex radios require effective self‑interference cancellation, and digital cancellation is a key step to remove residual interference. The paper proposes a novel digital self‑interference cancellation method that improves upon prior art. The method uses an adaptive RLS algorithm that adjusts the digital filter length according to the self‑interference channel. Experiments demonstrate the method achieves ~26 dB cancellation for a 5 MHz signal at 15 dBm transmit power and remains effective even when passive and active cancellation are imperfect, outperforming previous approaches.

Abstract

Previous research shows that it is possible to achieve full-duplex system as long as self-interference is cancelled enough. Digital cancellation method is an important step to remove the residue self-interference signal in digital domain. Compared with prior art, this paper purposes a novel digital cancellation method. The recursive least squares (RLS) algorithm is introduced and digital filter length can be adjusted adaptively based on the self-interference channel characteristic. The real-time experimental results show that our method has fairly desirable performance than previous research. It can achieve about 26dB for 5MHz bandwidth signal at high transmit power 15dBm. Furthermore, it is still effective when passive and active cancellation mechanisms are not ideal.