Enhanced Channel Estimation for Hybrid-Field XL-MIMO Systems Using Joint Sparse Bayesian Learning

Accurate channel estimation in hybrid-field extremely large-scale multiple-input multiple-output (XL-MIMO) systems is crucial for unlocking their potential in future wireless communications. Traditional methods, such as least squares (LS) and orthogonal matching pursuit (OMP), are limited by their inability to fully exploit correlations between near-field and far-field paths. To address these limitations, this letter first explores hybrid-field sparse Bayesian learning (Hybrid SBL), which separately applies the traditional SBL framework to near-field and far-field paths, and then develops a joint sparse Bayesian learning (Joint SBL) framework. By integrating near-field and far-field dictionaries into a unified structure and leveraging a global sparse prior, Joint SBL eliminates the need for prior path information while enhancing channel estimation accuracy and robustness. Simulation results illustrate that the proposed Joint SBL significantly outperforms traditional methods, especially under challenging signal-to-noise ratio (SNR) conditions.