Chapter Summary

1.
Product path loss ( $d_1^2 d_2^2$ ) makes optimal RIS placement near an endpoint (BS or UE), NOT at the geometric midpoint — the opposite of naive intuition
2.
In multi-UE deployments, near-UE placement dominates empirically — RIS near a UE cluster (mall entrance, street corner) sees the most traffic
3.
At mmWave, $p_{\text{blk}}$ can exceed 0.5; RIS coverage scales as $p_{\text{cov}}^{\text{total}} = p_{\text{LOS}}(1-p_{\text{blk}}) + p_{\text{RIS}} p_{\text{blk}}$ , typically boosting coverage from 40% to 85+%
4.
PPP-model coverage vs. density: $p_{\text{cov}}(\lambda) = 1 - e^{-\lambda \pi R^2}$ ; for urban 28 GHz with $R \approx 80$ m, $\sim 150$ panels/km² gives 95% coverage
5.
RIS beats AF relay on total cost when $N \geq \sqrt{N_t \cdot G_{\text{amp}} \cdot \eta_{\text{relay}}}$ , typically $N \geq 16$ — essentially always
6.
10-year TCO: RIS is $\sim 36\%$ of small-cell, $\sim 53\%$ of relay for coverage fill-in scenarios; the advantage grows with scale
7.
Cost-per-QoS $\Psi$ is the operator's deployment decision metric; greedy submodular placement gives $(1-1/e) \approx 63\%$ optimality guarantee
8.
RIS is a coverage tool, NOT a capacity tool; hybrid deployments with small cells for capacity and RIS for coverage fill-in are optimal
9.
The CommIT deployment framework (Caire-Atzeni 2023) achieves 95% coverage at 35% lower cost than facility-location baselines via joint deployment and network-level optimization