TL;DR
This paper introduces a novel Hybrid NOMA (H-NOMA) design that combines Hybrid Successive Interference Cancellation (HSIC) with Power Adaptation (PA). The proposed scheme achieves a "win-win" scenario: it provides higher data rates than traditional Orthogonal Multiple Access (OMA) while consuming less energy, and unlike previous models, it is mathematically guaranteed to outperform OMA at high signal-to-noise ratios (SNR) without restrictive constraints on user power or target rates.
What Problem Does This Paper Address?
As we move toward 6G, traditional OMA (where users get their own dedicated time/frequency slots) is struggling to keep up with massive connectivity demands. Non-Orthogonal Multiple Access (NOMA) allows multiple users to share the same resource by multiplexing them in the power domain. However, jumping straight to NOMA is difficult because existing infrastructure is built for OMA.
Hybrid NOMA (H-NOMA) bridges this gap by allowing "opportunistic" users to transmit during the time slots of "legacy" OMA users. The challenge is ensuring that the opportunistic user doesn't ruin the connection for the legacy user. Previous H-NOMA designs often hit a performance ceiling (an "error floor") or only worked under very specific, restrictive conditions regarding channel gains and target data rates. The authors set out to design a scheme that is robust, energy-efficient, and guaranteed to work better than OMA in high-interference environments.
Methods
The authors propose the HSIC-PA aided H-NOMA scheme for uplink transmissions. The core components are:
- Hybrid SIC (HSIC): Instead of a fixed decoding order at the base station, the system dynamically chooses the best order to decode signals based on real-time channel conditions.
- Power Adaptation (PA): This is the "secret sauce." The opportunistic user ($U_n$) doesn't just blast at a fixed power. Instead, it uses a power adaptation coefficient ($\gamma$) to adjust its signal. If the interference to the legacy user ($U_m$) is too high, $U_n$ scales back its power just enough to meet a specific interference threshold ($\tau_m$). This ensures the legacy user’s performance remains identical to OMA while allowing the opportunistic user to squeeze in extra data.
- Mathematical Framework: The authors derived rigorous closed-form expressions for the probability that their scheme outperforms OMA. They also performed asymptotic analysis to see how the system behaves as SNR approaches infinity.
Datasets
As a theoretical communication theory paper, no external real-world datasets were used. Instead, the authors evaluated their design using:
- Monte Carlo Simulations: To verify the accuracy of their derived closed-form expressions.
- Channel Models: Small-scale Rayleigh fading models were used to simulate typical wireless environments.
- Benchmarks: The proposed HSIC-PA scheme was compared against pure OMA, Fixed SIC (FSIC) H-NOMA, and HSIC without Power Adaptation (HSIC-NPA).
Key Insights and Ideas
- Breaking the Error Floor: Traditional NOMA schemes often suffer from "outage probability floors" at high SNR because interference becomes unmanageable. By introducing Power Adaptation, this paper shows that these floors can be completely eliminated.
- Deterministic Superiority: The most striking finding is that at high SNR, the probability of the proposed scheme outperforming OMA approaches 1 (100%). This happens regardless of the users' target rates or power ratios, which were major stumbling blocks in previous research.
- The Dominance of the Opportunistic User: The analysis reveals that the channel gain order of the opportunistic user ($n$) has a much larger impact on performance than the legacy user ($m$). In practical terms, this means the system should prioritize pairing legacy users with opportunistic users who have significantly better channel conditions.
Key Results
- Energy Efficiency: The proposed scheme is proven to consume less energy than OMA (specifically, a fraction $1+\gamma$ of the power, where $\gamma \leq 1$ and the total power is reduced by a coefficient $\beta < 1/2$).
- Rate Gains: Despite using less energy, the scheme achieves a higher cumulative data rate than OMA with a probability that converges to 1 as SNR increases.
- Numerical Validation: Simulations showed that while previous HSIC-NPA schemes plateaued (staying at a failure probability of $10^{-1}$ or $10^{-2}$), the proposed HSIC-PA scheme's failure probability continued to drop exponentially as SNR increased, reaching $10^{-4}$ and beyond.
Limitations and Future Work
- CSI Requirements: The model assumes perfect Channel State Information (CSI) at both the transmitter and receiver. In real-world "messy" environments, CSI is often imperfect, which could degrade SIC performance.
- Channel Dynamics: The study focuses on slow time-varying channels. The authors acknowledge that extending this to fast-fading scenarios where the channel changes within a single frame is a necessary next step.
- User Pairing: While the paper analyzes the impact of user ordering, it doesn't propose an optimized algorithm for pairing $M$ users into $m/n$ groups, which would be critical for a large-scale network deployment.
Summary updated