Visibility-Region Coupling in XL-MIMO AGV Fleets: Triple-Role Modeling and Masked Beamforming

Extremely large-scale multiple-input multiple-output (XL-MIMO) is a promising technology for supporting automated guided vehicle (AGV) fleets in smart port terminals. However, the metallic container environment induces spatial non-stationarity, whereby each AGV is visible to only a subset of the array, referred to as its visibility region (VR). Unlike existing XL-MIMO models that assume user-independent VRs, we show that each AGV simultaneously acts as a communication user, a metallic scatterer, and a blocker, resulting in coupled user channels and VRs. We formulate this \emph{triple-role} effect through a VR-coupled channel model and develop a VR-aware downlink beamforming framework based on masked weighted minimum mean-square error (WMMSE), where the masking operation exactly enforces VR support constraints while significantly reducing computational complexity. Simulation results in a realistic smart port scenario demonstrate more than a threefold sum-rate improvement over VR-unaware baselines, with the gains becoming increasingly pronounced as fleet density increases.