Adaptive Mode-Switching STAR–IRS-Assisted Uplink NOMA Communication Systems

We propose a simple yet effective method for selecting the number of simultaneous transmission and reflection (STAR-IRS) elements in a NOMA uplink with one reflection and one transmission user under a mode-switching (MS) scheme. Unlike prior works that rely on fixed element allocation, our scheme dynamically assigns elements based on instantaneous channel conditions using a short pre-transmission phase with NOMA test symbols and 1-bit feedback. Closed-form expressions for outage probability and sum throughput are derived and validated through simulations. Results show throughput gains of up to 28%, 45%, and 78% at SNRs of 30, 16, and 8 dB for 4, 8, and 16 IRS elements, respectively, compared with optimal fixed allocation. Furthermore, the scheme requires minimal IRS element utilization, thereby reducing channel estimation overhead at the IRS controller as the estimation takes place on the fly. At moderate and higher SNR levels, the system achieves superior throughput performance while requiring fewer average number active IRS elements.