Performance of IRS-Assisted NOMA Communication with Dynamic IRS Allocation

This paper investigates a novel yet simple method for selecting the number of intelligent reflecting surface (IRS) elements in the IRS-assisted non-orthogonal multiple access (NOMA) scheme to serve both near and far users. Unlike conventional IRS selection, which maximizes throughput with a fixed number of selected IRS elements for all transmission times, the proposed scheme adapts the number of selected IRS elements dynamically based on the channel conditions between the source and the IRS, the IRS and the near user, and the IRS and the far user. Closed-form expressions for outage probability and achieved throughput are derived for both the near and far users. In comparison to the conventional statically optimized IRS selection method, the proposed method achieves significantly better throughput gains with the same number of IRS elements considered. All the derived results are validated through extensive computer simulations.