Joint Dictionary Learning and Channel Estimation in Hybrid-Field XL-MIMO Systems

Hybrid-field channel estimation in extremely large-scale multiple-input and multiple-output systems presents significant challenges due to the coexistence of far-field and near-field scatterers. This letter proposes an alternating direction method of multipliers-based joint dictionary learning and channel estimation method to address these challenges. By leveraging extensive channel measurement data, a hybrid-field dictionary is learned to effectively capture the unique characteristics of the hybrid-field channels. Simulation results demonstrate that our proposed method outperforms existing predefined far-field and near-field dictionary-based schemes, highlighting its potential for Internet of Things networks.