Measurement-Efficient Port-Reading Policies for Fluid Antenna Systems Under Spatially Correlated κ-μ Shadowed Fading

Fluid antenna systems exploit spatial diversity by selecting among multiple candidate ports, but practical receivers may be unable to read all ports before selection. This work studies the port-reading problem, namely, which ports should be measured under a strict observation budget. Under spatially correlated κ-μ shadowed fading, geometry-driven uniform reading, threshold-aware marginal ranking, and greedy coverage are compared in terms of observed-only outage probability. For an aperture with 100 candidate ports and observation budgets of 3 to 7 measured ports, the results show a clear budget-dependent behavior. Greedy coverage is most effective with three reads, achieving the lowest outage in 12 of 14 fading regimes. For four or more measured ports, uniform subsampling becomes the most reliable strategy and achieves the best average outage across regimes. Marginal ranking consistently underperforms, indicating that per-port success frequency is a poor proxy for set-level diversity in correlated apertures. These results establish port reading as a critical design layer for measurement-efficient operation of fluid antennas.