A Convex Demixing Approach for Hybrid-Field Channel Estimation of XL-MIMO Systems via Atomic Norm Minimization

Channel estimation is a critical task in extremely large-scale multiple-input multiple-output (XL-MIMO) systems for 6G wireless communications. A hybrid-field channel model effectively characterizes the mixed far-field and near-field scattering components in practical XL-MIMO systems. In this paper, we propose a convex demixing approach for hybrid-field channel estimation within the atomic norm minimization (ANM) framework. By promoting sparsity of the far-field and near-field components directly in the continuous parameter domain, a demixing scheme that minimizes a weighted sum of two atomic norms is proposed. We show that the resulting ANM is equivalent to a computationally feasible semi-definite programming (SDP). Numerical experiments on simulated data demonstrate that our method outperforms existing approaches for hybrid-field channel estimation.