Hybrid IRS-assisted downlink NOMA: a deployment strategy for future wireless networks

Intelligent Reflecting Surfaces (IRS) have emerged as a promising paradigm for reshaping wireless propagation environments and boosting network performance. This paper investigates the performance enhancement of non-orthogonal multiple access (NOMA) systems through a hybrid IRS architecture in which reflecting elements are jointly deployed on an unmanned aerial vehicle (UAV) and a building facade. Three deployment strategies are examined: a UAV-mounted IRS (UIRS), a building-mounted IRS (BIRS), and the proposed hybrid configuration where near users are served via the BIRS while far users benefit from the superior line-of-sight (LoS) links provided by the UIRS. While the UIRS offers enhanced coverage at the cost of higher energy consumption, the BIRS provides a power-efficient yet spatially limited alternative. Our objective is to maximize the system sum rate under this hybrid IRS-assisted NOMA framework. The resulting optimization problem—jointly involving IRS phase shift design and power allocation at base-station is inherently non-convex; semidefinite relaxation (SDR) is employed to obtain the near-optimal solutions. Numerical simulations demonstrate that the proposed hybrid IRS model consistently outperforms both standalone UIRS and BIRS schemes in terms of achievable sum rate, underscoring its potential for next-generation 5G-and-beyond wireless networks.