Adaptive Cancellation of Nonlinear Self-Interference in Wireless Full-Duplex: Cascaded Spline-Interpolated Methods

Abstract

<p>This article studies advanced digital self-interference (SI) cancellation algorithms for wireless full-duplex (FD) transceivers. While the majority of works in this area consider linear SI models and corresponding cancellation algorithms, we take also both the transmitter (TX) and receiver (RX) nonlinearities into account. Specifically, we develop a Hammerstein-Wiener type of digital SI cancellation system where the two involved nonlinearities are efficiently implemented through spline-interpolated look-up tables (LUTs). Also, adaptive parameter estimation solutions are pursued and discussed, and comprehensive cancellation performance examples and processing complexity comparisons are provided. The obtained results show that taking both the TX and RX nonlinearities into account in the digital cancellation stage can be beneficial, particularly when the TX power amplifier is operating close to saturation and when the RF isolation and RF cancellation prior to the RX low-noise amplifier (LNA) is limited, such that also the LNA is distorting the SI waveform.</p>