Intelligent Reflecting Surface Joint Uplink-Downlink Optimization for NOMA Network

Abstract

This paper investigates the performance of joint uplink-downlink communication of intelligent reflecting surface (IRS) assisted non-orthogonal multiple access (NOMA) network. Unlike most existing works that considered time-division duplexing (TDD) system to exploit the IRS uplink-downlink channel reciprocity, we adopt frequency-division duplexing (FDD) to achieve a fair trade-off between the signal reception reliability and system spectral efficiency. We analyze the system outage probability and outage-throughput, and derive their associated performance bounds in closed-form expressions. Moreover, for the second user in decoding order, we formulate two optimization problems over the IRS elements phase-shifts. The first optimization problem aims to maximize the minimum SNR in the uplink and downlink and the second optimization problem targets maximizing both SNRs. We employ genetic algorithms (GA) to solve the two problems. Monte-Carlo simulations are applied to validate the analytically driven bounds and to compare between the solutions of the proposed optimization problems.