Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces 3DCENet, a deep learning-based channel estimation network for MIMO systems that employs a dual attention mechanism to capture channel correlations across time, frequency, and spatial domains. By jointly considering these domains, 3DCENet avoids the performance degradation associated with suboptimal, decomposed estimation methods. Experiments demonstrate that 3DCENet achieves accurate channel estimates without the computational burden of full LMMSE, offering a practical alternative for correlated MIMO channels.
Achieve accurate MIMO channel estimation without the prohibitive complexity of LMMSE by jointly attending to time, frequency, and spatial channel correlations.
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.