Abstract
This paper investigates the performance of Unmanned Aerial Vehicle (UAV)-aided multi-way relaying networks (MWRNs) employing Non-Orthogonal Multiple Access (NOMA) while considering the impact of residual hardware impairments (RHIs). By leveraging NOMA, the system reduces the required communication time slots to just two, significantly enhancing the achievable sum-rate (ASR) compared to traditional orthogonal schemes. However, analytical results reveal that performance is ultimately constrained by an interference floor caused by RHIs and inter-user interference. We provide closed-form expressions for ASR and conduct high-SNR asymptotic analyses to characterize the system's behavior under realistic hardware conditions.
Introduction
As the industry moves toward 6G connectivity, two technologies have emerged as pivotal: Unmanned Aerial Vehicles (UAVs) and Non-Orthogonal Multiple Access (NOMA). UAVs offer flexible, line-of-sight (LoS) relaying capabilities, while NOMA allows multiple users to share the same frequency and time resources, drastically improving spectral efficiency.
In traditional Multi-Way Relaying Networks (MWRNs), multiple users exchange information through a central relay. When using Orthogonal Multiple Access (OMA), the number of time slots required for a full exchange scales with the number of users, creating a significant bottleneck. NOMA addresses this by compressing the exchange into fewer phases. However, most theoretical models assume "perfect" hardware. In reality, physical transceivers suffer from Residual Hardware Impairments (RHIs)—such as phase noise, I/Q imbalance, and amplifier non-linearities—that persist even after software compensation. This study tackles the challenge of modeling a high-speed UAV relay system that is both efficient and realistic by accounting for these impairments.
System Model and Methodology
The researchers modeled a system where $M$ ground users exchange information via an Amplify-and-Forward (AF) UAV relay. The communication process is divided into two distinct phases: the Multiple-Access (MAC) phase, where users transmit to the UAV, and the Broadcast (BC) phase, where the UAV relays the combined signal back to the users.
Key analytical techniques employed include:
- Nakagami-m Fading: This statistical model describes the wireless channels more generally and realistically than standard Rayleigh fading, accounting for varying degrees of LoS and scattering.
- RHI Modeling: Distortion noise terms were incorporated at both the transmitter and receiver ends of every node to simulate hardware imperfections.
- Mathematical Derivation: The core contribution involves deriving closed-form analytical expressions for the Achievable Sum-Rate (ASR).
- Asymptotic Analysis: A "High SNR" analysis was performed to determine system behavior at high power levels, specifically identifying the "high SNR slope" and "power offset."
To validate these mathematical formulas, the authors utilized Monte Carlo simulations with $10^5$ independent trials, ensuring the theoretical derivations align with simulated performance.
Key Insights and Results
The Efficiency of NOMA
In traditional OMA systems, the time required for information exchange grows linearly with the number of users. By implementing NOMA, the authors demonstrate that the entire exchange can be compressed into just two time slots, regardless of the number of users. This leads to a significantly higher ASR, particularly as the network scales.
The Interference Floor
A critical finding in the high-SNR analysis is the existence of a performance "ceiling." Even with infinite transmit power, the ASR does not increase indefinitely. This is because the interference from other users and the hardware distortion noise scale proportionally with the signal power, creating a constant signal-to-interference-plus-noise ratio (SINR) at high power levels.
Impact of Hardware Impairments
The study reveals that RHIs at the transmitter and receiver have nearly identical detrimental effects on the total sum-rate. The performance gap between ideal hardware and realistic hardware becomes more pronounced as the number of users increases, highlighting the necessity of considering RHIs in 6G system design.
UAV Positioning
Results confirm that ASR is highly sensitive to the UAV's spatial coordinates. Moving the UAV to a central location that minimizes the average distance to all ground users significantly optimizes performance. This underscores the advantage of UAV mobility over fixed relay infrastructure.
Conclusion and Future Work
While NOMA-enhanced UAV networks show great promise, several constraints remain. This study assumed a half-duplex constraint, where nodes cannot transmit and receive simultaneously. Future research into Full-Duplex operation could potentially double efficiency, though it introduces complex self-interference issues. Additionally, the authors suggest that joint optimization of UAV placement and power allocation strategies could further enhance the performance of these networks in the presence of hardware imperfections.
Summary updated