Chapter Summary

Chapter Summary

Key Points

• 1.

  STAR-RIS serves the full space. Each element produces a reflected wave (to the incident-side) and a transmitted wave (to the other side), characterized by $r_n$ and $t_n$ respectively. Passivity enforces $|r_n|^2 + |t_n|^2 \leq 1$. Hardware coupling typically constrains $\theta^t - \theta^r = \pm\pi/2$ unless active/chiral metasurfaces are used.

• 2.

  Three protocols: ES, MS, TS. Energy Splitting (continuous amplitude split per element), Mode Switching (binary per-element reflect/transmit), Time Switching (homogeneous alternating across time). Performance: ES $\geq$ MS, ES $\geq$ TS; MS and TS incomparable in general but MS approaches ES at large $N$.

• 3.

  Coverage gain of $\sim 1.5$-$2\times$ over passive RIS. The geometric doubling (reflect + transmit sides) is slightly offset by the per-side energy split. Deployments where behind-the-panel coverage is required (indoor-outdoor, dense 6G, V2X) benefit most. For single-side deployments, passive RIS is simpler and faster.

• 4.

  AO extends naturally: three-block update. Active precoder update, reflection-side $\boldsymbol{\Phi}^r$ update, transmission-side $\boldsymbol{\Phi}^t$ update, and amplitude reallocation (ES). Monotone convergence; similar iteration count to passive RIS. Per-iteration cost $\sim 1.5$-$2\times$ passive RIS.

• 5.

  Protocol choice is hardware-driven. ES requires continuous amplitude control; MS needs only switches; TS reuses passive-RIS hardware. Commercial 2024 panels lean MS with 3-bit phases for a good balance of performance and simplicity.