Blind LMS Algorithm with Multiple Error Cost Function for Self-Interference Compensation in Full-Duplex Relays

G. GONZÁLEZ; F. GREGORIO; J. E. COUSSEAU

Cordoba

Congreso; XVI Reunión de Trabajo en Procesamiento de la Información y Control; 2015

Full-duplex relays will play a key role in futurenetworks since they increase the system diversity, coverage, andthroughput; ideally doubling the spectral efficiency. However, thefull duplex operation produces coupling between the transmitterand receiver front ends that drastically reduces the relay performanceand it can even turn it unstable. Then, to allow a fullduplex operation it is necessary to compensate for the coupling.The autocorrelation pattern of the baseband received signal isperturbed by the relay self-interference. We propose a blindadaptive least mean squared (LMS) algorithm to compensate forthe coupling that employs the known second-order statistics of thesource signal as a reference. In order to improve the cancelerperformance, we propose a novel multiple error cost functionthat includes the information of several autocorrelation lags toadapt the canceler coefficients. Additionally, we present a studyof the proposed criterion and analyze the stationary points of theadaptive algorithm. Finally, we show the canceler performanceusing simulations.