Feedback-controlled Channel Estimation with Low-resolution Adcs in Multiuser Mimo Systems

Rodrigo Pereira David; João Cal-Braz

ICASSP 2019accepted0 citations

Feedback-controlled Channel Estimation with Low-resolution Adcs in Multiuser Mimo Systems

Abstract

The nonlinear nature of the quantization operation performed by the ADCs imposes great challenges to the systems employing coarse quantization. Systems operating in moderate to high SNR regimes experience a pronounced capacity loss and also impairments in the quality of the channel estimates, compared to the infinite precision case. This work proposes a multiuser MIMO channel estimation algorithm that operates with low-resolution receiving ADCs. The strategy is based on an coarse quantizer followed by an EM-ML estimator, that takes as input the observation vector subtracted by an offset value produced by a first rough channel estimate. The approach performed by this feedback-controlled double-stage estimator aims to make the quantizer thresholds based on the low-SNR Cramér-Rao lower bound approximation valid in the high SNR region. Numerical analysis reveals that reliable estimation is achieved in coarse quantization for a wide range of SNR with a very small and constant gap in relation to infinite precision quantized estimation.

BibTeX

@inproceedings{icassp2019_feedbackcontroll,
  title = {Feedback-controlled Channel Estimation with Low-resolution Adcs in Multiuser Mimo Systems},
  author = {Rodrigo Pereira David and João Cal-Braz},
  booktitle = {ICASSP 2019},
  year = {2019}
}