Data Augmentation Empowered Neural Precoding for Multiuser MIMO with MMSE Model

TL;DR

This paper introduces iPNet, an interpretable neural precoding network for multiuser MIMO systems that mimics MMSE precoding, leveraging data augmentation to improve performance and robustness over traditional black-box neural networks.

Contribution

The paper proposes an interpretable neural precoding architecture, iPNet, that combines model-driven and data-driven components to enhance performance and generalizability in MU-MIMO systems.

Findings

iPNet outperforms existing black-box neural precoding methods.

Augmented CSI improves precoding accuracy.

iPNet shows robustness against CSI mismatches.

Abstract

Precoding design exploiting deep learning methods has been widely studied for multiuser multiple-input multiple-output (MU-MIMO) systems. However, conventional neural precoding design applies black-box-based neural networks which are less interpretable. In this paper, we propose a deep learning-based precoding method based on an interpretable design of a neural precoding network, namely iPNet. In particular, the iPNet mimics the classic minimum mean-squared error (MMSE) precoding and approximates the matrix inversion in the design of the neural network architecture. Specifically, the proposed iPNet consists of a model-driven component network, responsible for augmenting the input channel state information (CSI), and a data-driven sub-network, responsible for precoding calculation from this augmented CSI. The latter data-driven module is explicitly interpreted as an unsupervised learner of the MMSE precoder. Simulation results show that by exploiting the augmented CSI, the proposed iPNet achieves noticeable performance gain over existing black-box designs and also exhibits enhanced generalizability against CSI mismatches.