Application of ICA on Self-Interference Cancellation of In-Band Full Duplex Systems

Abstract

In this letter, we propose a modified version of Fast Independent Component Analysis (FICA) algorithm to solve the self-interference cancellation (SIC) problem in In-band Full Duplex (IBFD) communication systems. The complex mixing problem is mathematically formulated to suit the real-valued blind source separation (BSS) algorithms. In addition, we propose a method to estimate the ambiguity factors associated with ICA lumped together with the channels and residual separation error. Experiments were performed on an FD platform where FICA-based BSS was applied for SIC in the frequency domain. Experimental results show superior performance compared to least squares SIC by up to 6 dB gain in the SNR.

Blind Source Separation For Full-Duplex Systems: Potential and Challenges

“…Table I summarizes the parameters used in our framework. More details of the experimental framework have been discussed in [9], [41].…”

“…Thus, a complex-valued ICA has to be deployed for suppression. However, matrix decomposition must be applied to convert the complex-valued mixing matrix to a real-valued matrix, such that real-valued ICA can be used, as performed in [9].…”

“…where Y si and Y soi are the separated SI and SOI signals, respectively; β is the scaling ambiguity for the recovered selfinterference signals; and W is the demixing matrix. It is worth mentioning that the residual separation error (as discussed in [37]) is ignored in this paper. This residual separation error is due to the imperfect separation of the ICA due to reaching stopping criterion and the limited frame length.…”

“…Ambiguity and permutation problem: BSS techniques are always associated with ambiguity, which requires the use of estimation techniques, such as pilot-assist techniques. For instance, non-overlapped long preambles have been used to assist in solving this problem in [9]. However, such a solution might be not preferable, due to incompatibility with existing standards.…”

“…Recently, it has been proposed in [8], [9] that the independent component analysis (ICA) approach, which was originally employed as a blind source separation (BSS) technique, can be applied to separate the mixed signals in a full-duplex system. The system setup and the modified Fast ICA algorithm, based on the real domain, have been presented in [9]. Experimental results have shown that the performance can be improved by up to 6 dB, compared to the least-squares interference cancellation method.…”