A New Online Nonlinear Self-Interference Cancelation Method With Random Fourier Features

In this letter, we present a new online nonlinear digital self-interference cancelation method for full duplex radios. In this method, random Fourier features transformation is performed before utilizing the least mean squares algorithm, which is applied to the residual of linear digital cancelation. The proposed solution's performance is evaluated against prominent nonlinear cancelation schemes on the same full duplex radio set-up, consisting only of a monostatic antenna and a WARP v3 board. Our experimental results show that the proposed solution provides the highest self-interference cancelation while reducing the complexity of the batch solutions by up to two orders of magnitude.