A Comparative Study of Digital Self-interference Cancellation Techniques in In-Band Full-Duplex OFDM Systems

Abstract

In-Band Full-Duplex is a powerful technique which theoretically doubles the spectral efficiency. It significantly contributes to the ever-increasing demand for high data rate and good connectivity of future communication systems. Self-Interference from the transceiver is the most challenging issue in In-Band Full-Duplex system implementation. In the presence of transmitter power amplifier non-linearity and phase noise in local oscillator, cancellation of the self-interference becomes more complex. In this paper, a detailed comparative study of two major Digital Self-Interference Cancellation techniques namely, Least Squares method based Dynamic Regression for Nonlinear Digital Self-Interference Cancellation and Least Mean Squares based Adaptive Nonlinear Digital Self Interference Cancellation is done. The study of the algorithms is done with Orthogonal Frequency Division Multiplexing that uses a bit-interleaved coded modulation system. The performance of the algorithms is studied in terms of bit error rate, convergence time and computation complexity in the presence of time varying fading channel with power amplifier non-linearity and phase noise.