Investigation of NOMA 5G Systems Under Non-Gaussian Channels
Hasan Abu Hilal
DOI: https://doi.org/10.1007/s40031-023-00955-3
2023-12-03
Journal of The Institution of Engineers (India) Series B
Abstract:This research investigation centers around the comprehensive analysis of the operational efficiency of nonorthogonal multiple-access (NOMA) technology within the presence of impulsive noise disturbances. This technological paradigm is of paramount importance for the trajectory of wireless communication advancements and warrants meticulous scrutiny. While the present global populace predominantly relies on 4G networks, the impending launch of 5G networks prompts the incorporation of NOMA to optimize its prospective deployment. However, wireless communication mediums grapple with a spectrum of challenges that directly influence their operational efficacy, with noise disruptions constituting a pivotal concern. While the conventional literature often presupposes the prevalence of Gaussian noise, this study delves into the domain of non-Gaussian noise, characterized by divergent ramifications for system dynamics. Notably, the integration of multiuser detection mechanisms into wireless channels in the context of non-Gaussian noise has generated scholarly skepticism. The principal aim of this empirical inquiry is to discern the discernible impact of non-Gaussian noise on distinct NOMA detection methodologies. To achieve this objective, a detector grounded in the principles of compressive sensing is harnessed, leveraging the latent architectural inadequacies intrinsic to NOMA systems. The research endeavor encompasses a gamut of noise parameter configurations, each explicitly exposing the diverse implications of noise on the bit error rate (BER) profiles of NOMA configurations. Additionally, a controlled experimental framework is deployed, juxtaposing BER performances attained from NOMA systems against those originating from Rayleigh flat fading channels. The BER performance is rigorously scrutinized across a spectrum of conditions, spanning iterative support detection, structured iterative support detection, and orthogonal matching pursuit. As the signal-to-noise ratio (SNR) scales ascend, discernible shifts transpire within the gradient of the performance profiles, indicative of impinging errors interacting with intrinsic vulnerabilities. Amidst this dynamic, the BER incurred from non-Gaussian noise exhibits an idiosyncratic trajectory, deviating from the typical trends governed by additive white Gaussian noise (AWGN). Evident from the analysis, BER exhibits a descending trend commensurate with amplifying SNR, albeit plateauing once a distinct SNR threshold is attained. This temporal profile initially demonstrates sanguine behavior, yet it subsequently undergoes degradation with the crescendo of SNR values, attributed to disruptive interference stemming from the architectural frailty. The BER manifestations traverse an intricate spectrum contingent upon noise levels and the specific modality of detection in implementation. Singularly attributing superiority to any one method proves to be a formidable endeavor, given the nuanced interplay of noise and detection mechanisms.