RF interference cancellation based on time-varying step-size least mean square algorithm

A new cancellation method based on time-varying step-size least mean square algorithm is proposed to compromise the convergence speed and cancellation ratio of radio frequency (RF) cancellation methods for co-frequency and co-time full duplex (CCFD) in this paper. This method is constructed on cancellation scheme with a composite of in-phase and quadrature signals, and the nonlinear characteristic of logistic function is utilized to vary step-size. Then the two attenuators are accurately fine-tuned by varying step-size and least mean square criterion to cancel the self-interference. Finally, we analyze the convergence and misadjustment steady-state error of this interference cancellation method in theory, and extract the closed expression of convergent cancellation ratio. Theory analysis and simulations verify that convergence time of proposed method is only a quarter of existing methods and lasted cancellation ratio is more 5 dB than existing methods.