Dual-Attention Based 3D Channel Estimation

TL;DR

This paper introduces 3DCENet, a deep learning model utilizing dual attention mechanisms to improve 3D channel estimation in MIMO systems, addressing complexity and correlation challenges.

Contribution

The paper proposes a novel dual attention based deep learning network for 3D channel estimation, enhancing accuracy over traditional suboptimal methods.

Findings

3DCENet achieves higher estimation accuracy than conventional methods.

The dual attention mechanism effectively captures correlations across domains.

The approach reduces complexity compared to full 3D LMMSE filtering.

Abstract

For multi-input and multi-output (MIMO) channels, the optimal channel estimation (CE) based on linear minimum mean square error (LMMSE) requires three-dimensional (3D) filtering. However, the complexity is often prohibitive due to large matrix dimensions. Suboptimal estimators approximate 3DCE by decomposing it into time, frequency, and spatial domains, while yields noticeable performance degradation under correlated MIMO channels. On the other hand, recent advances in deep learning (DL) can explore channel correlations in all domains via attention mechanisms. Building on this capability, we propose a dual attention mechanism based 3DCE network (3DCENet) that can achieve accurate estimates.