BER Analysis of Downlink NOMA over AWGN, Rayleigh, and Rician Channels under Imperfect SIC with Fairness-Aware Power Allocation and OMA Baseline

Abstract

This paper investigates the bit-error-rate (BER) performance of a two-user downlink power-domain non-orthogonal multiple access (PD-NOMA) system under AWGN, Rayleigh, and Rician fading. Unlike baseline studies that assume ideal successive interference cancellation (SIC), the proposed framework explicitly models imperfect SIC through a residual interference factor beta and quantifies its effect on near-user reliability. To address reviewer concerns regarding methodological rigor, the manuscript now includes: i) a direct orthogonal multiple access (OMA) baseline comparison under the same channel conditions and transmit-power normalization; ii) corrected and fully harmonized simulation parameters; and iii) a strengthened novelty positioning against recent BER- and fairness-oriented NOMA literature. Monte Carlo results show that AWGN provides the lowest BER, Rayleigh fading yields the largest BER due to deep fades, and larger Rician K-factors improve low-to-medium-SNR performance. Under imperfect SIC, the near user exhibits an interference-limited error floor at high SNR. A fairness-aware grid-search power allocation reduces worst-user BER while preserving decoding order. The added OMA comparison clarifies the reliability-efficiency trade-off: OMA generally provides lower BER, whereas NOMA retains the spectral-efficiency advantage but becomes increasingly sensitive to SIC quality. The resulting framework offers a reproducible and practically grounded BER benchmark for mixed LOS/NLOS NOMA deployments.