Abstract
Independent component analysis (ICA) has been shown as a promising means for solving the self-interference cancellation (SIC) problem for in-band full duplex systems. This paper presents a detailed analysis of the interference suppression capability and computational complexity of several different ICA algorithms, operating at either real-valued or complex-valued domain. In addition, on the basis of the setup of the full-duplex system, we show that a much simplified complex-valued ICA algorithm that only performs whitening and decorrelation processes are sufficient to separate the signal of interest from the mixed signal. Extensive simulation results are presented in this paper to illustrate the performance and complexity of various ICA approaches applying to the full-duplex system.