Energy Efficiency Maximization for Active IRS-Assisted Uplink NOMA Systems

Abstract

Active intelligent reflecting surface (IRS) can be coupled by non-orthogonal multiple access (NOMA) to enhance the spectrum efficiency. However, this improvement comes at the expense of increased power consumption, which may lead to a decline in energy efficiency (EE). This motivates us to explore the EE performance in an active IRS-assisted uplink NOMA network. In the proposed approach, multiple users are assumed to transmit information to the base station (BS) aided by active IRS. We first develop a novel user ordering based on the strength of cascaded channels. Then, we maximize the EE by properly designing the transmit power and IRS active beamforming. This non-convex problem is addressed with an effective algorithm involving semi-definite relaxation and alternating optimization. Simulation results validate that the presented algorithm can provide marked EE improvement and outperform the benchmarks.