Data-Aided MU-MIMO Channel Estimation Utilizing Gaussian Mixture Models

Franz Weißer, Dominik Semmler, Nurettin Turan, Wolfgang Utschick

Abstract

This work extends two previously proposed semi-blind channel estimators to a more general multi-user multiple-input-multiple-output (MU-MIMO) system. These estimators utilize data symbols in addition to pilot symbols to enhance the channel estimation quality. Based on all received signals, a subspace is calculated, which enhances the Gaussian mixture model based channel estimator. To estimate this subspace, we consider the inherent additional degrees of freedom in terms of precoding in MU-MIMO systems. Numerical simulations for different scenarios show that the extended methods outperform the studied state-of-the-art channel estimators.
BibTeX
@inproceedings{weier2024data,
  author = {Franz Weißer and Dominik Semmler and Nurettin Turan and Wolfgang Utschick},
  title = {{Data-Aided MU-MIMO Channel Estimation Utilizing Gaussian Mixture Models}},
  booktitle = {ICC},
  year = {2024},
  doi = {10.1109/icc51166.2024.10622811},
}

Loading related content...

Figure