Channel Estimation for Multiuser Extremely Large-Scale MIMO Systems

Existing channel estimation algorithms for ex-tremely large-scale multiple-input multiple-output (XL-MIMO) systems are predominantly designed for single-user scenarios and often overlook inter-user correlations. To address this limitation, this paper reformulates the joint multiuser channel estimation problem as a multiple-measurement vector (MMV)-based sparse signal recovery task. To solve this, we propose a novel row-wise hierarchical prior model that captures the structured sparsity of the joint multiuser channel in the angular-delay domain. Specifically, shared precision parameters for each row of the angular-delay domain channel are introduced to model common-row sparsity, while a Markov random field (MRF) is employed to encourage cluster sparsity. Building on this structured prior, we develop a computationally effi-cient channel estimation algorithm using variational message passing. Simulation results demonstrate that the proposed method significantly outperforms existing single-user-based approaches.