"Occupancy-Robust Layout--Material Co-Design for Wireless-Friendly Buildings dataset and code"

"This dataset and code develops the first BWP-side analytical co-design framework that takes occupancy uncertainty as a primary design input rather than a diagnostic concern. Human blockers are modeled as a Poisson point process, and closed-form expressions are derived for the first- and second-order channel-gain statistics, allowing the variance to enter a single mean--variance robust metric directly rather than through post-hoc Monte Carlo.  Results show that a static-optimal design inflates the channel-gain variance by roughly twofold at moderate occupancy, whereas the proposed robust design halves the variance over the entire occupancy range at the cost of a few percent in the mean-rate surrogate, and migrates from the strongest-reflection boundary to a clearly detuned multilayer configuration. "