Radio Frequency Self-Interference Cancellation With Analog Least Mean-Square Loop

A multitap adaptive filter with analog least mean-square (ALMS) loop is proposed in this paper for effective and low complexity self-interference cancellation implemented as part of the radio frequency frontend in a full duplex transceiver. Comprehensive analyses of the ALMS loop's behaviors at both micro and macroscales are presented for a wireless communication system with single carrier signaling. It is revealed that there is always an irreducible residual interference due to the cyclostationary property of the transmitted signal. The interference suppression ratio (ISR) lower bound is derived accordingly, which can be used as a design rule for determining the ALMS loop parameter. Stationary analysis shows that the convergence speed and achievable ISR of the ALMS loop are determined by the loop gain and the autocorrelation function of the transmitted signal. The interference channel modeling error with the adaptive filter also accounts for part of the residual interference power. These theoretical findings are verified by simulation and experimental results.