Iterative Semi-Blind CFO Estimation, SI Cancelation and Signal Detection for Full-Duplex Systems.
Yujie Liu,Xu Zhu,Eng Gee Lim,Yufei Jiang,Yi Huang
DOI: https://doi.org/10.1109/GLOCOM.2018.8647485
2018-01-01
Abstract:We propose an iterative semi-blind carrier frequency offset (CFO) estimation, self-interference (SI) cancelation and signal detection scheme for full-duplex (FD) orthogonal frequency division multiplexing (OFDM) systems. To the best of our knowledge, this is the first work to consider signal detection of FD systems in the presence of both CFO and SI. The CFO estimation, SI cancelation and signal detection are performed initially by a subspace based semi-blind method, which are then enhanced significantly by performing iterations among them. Its CFO compensation is performed on the desired signal estimate, avoiding the introduction of CFO to the SI. The pilots for the desired signal and SI are carefully designed to enable simultaneous transmission of them to achieve FD training mode. Simulation results show that, the proposed iterative scheme, with much lower training overhead, demonstrates a significant performance enhancement over the existing methods. By utilizing the second order statistics of the received signal, a much superior bit error rate (BER) performance can be achieved compared to the case with perfect SI cancelation and CFO compensation. Its output signal-to-interference-and-noise-ratio (SINR) is close to that with perfect SI cancelation, and robust against the input signal-to-interference ratio (SIR).