Abstract
As 5G networks evolve, balancing high overall capacity with equitable service for cell-edge users remains a critical challenge. This paper explores the synergy between Non-Orthogonal Multiple Access (NOMA) and Distributed Antenna Systems (DAS). By leveraging Remote Radio Units (RRUs) to specifically support weak users while a central base station serves the entire cell, the authors propose two transmission strategies: Single Selection and Blanket Transmission. Simulation results indicate that these NOMA-DAS configurations significantly outperform conventional NOMA in both max-min fairness and average sum-rate, especially as users move toward the cell periphery.
Introduction
The transition to 5G and beyond necessitates solutions for two persistent issues: maximizing total data throughput (capacity) and ensuring that users at the cell edge receive adequate signal quality (fairness). Distributed Antenna Systems (DAS) address this by distributing Remote Radio Units (RRUs) throughout the cell, effectively shortening the distance between the transmitter and the user. Concurrently, Non-Orthogonal Multiple Access (NOMA) enables multiple users to share spectral resources through power-domain multiplexing and Successive Interference Cancellation (SIC). Combining these technologies offers a promising path toward high-performance, fair wireless networks.
System Model and Proposed Methods
The research focuses on a single-cell environment comprising one central Base Station (BS) and six RRUs. The system serves two users: a "near" user located close to the BS and a "far" user positioned toward the cell edge.
Two primary transmission strategies are proposed:
- NOMA with Single Selection Transmission (SST): The central BS transmits to both users using NOMA. Only the RRU with the strongest channel to the far user provides additional data transmission to that user.
- NOMA with Blanket Transmission (BT): The central BS serves both users, while all RRUs simultaneously transmit the far user's data.
In these models, RRU signals are treated as beneficial boosts rather than interference. For the far user, this increases received signal power. For the near user, the additional power allocated to the far user's signal actually facilitates more accurate SIC, allowing the near user to better isolate and decode its own data.
Key Insights and Results
The performance of these strategies was evaluated using Monte Carlo simulations under various Channel State Information (CSI) scenarios, including instantaneous and statistical (CDI) knowledge.
Distance Resilience and Fairness
A standout finding is the "distance resilience" of NOMA-DAS. In conventional NOMA, performance typically degrades linearly as a user moves away from the central BS. However, in the proposed NOMA-DAS schemes, performance remains high and stable even as the far user approaches the cell edge, thanks to the proximity of the RRUs.
Blanket Transmission vs. Single Selection
The results consistently show that Blanket Transmission outperforms Single Selection. While coordinating multiple RRUs adds complexity, the resulting power gain for the weak user significantly enhances the overall system's max-min fairness.
Efficiency and Throughput
NOMA-DAS maintains superior total throughput (sum-rate) even under strict minimum rate requirements for the far user. While conventional NOMA often faces outages or significant performance drops when forced to prioritize fairness, the proposed schemes exhibit much lower performance loss.
Conclusion and Future Work
This study demonstrates that integrating NOMA with DAS provides a robust framework for achieving both high capacity and user fairness in 5G networks. By utilizing RRUs as SIC enablers, the proposed schemes mitigate the cell-edge problem effectively.
Future research directions include:
- Multi-User Scaling: Extending the framework beyond the two-user scenario to accommodate $K$ users.
- Power Optimization: Dynamically optimizing the power split between the central BS and the RRUs.
- Multi-Cell Coordination: Accounting for inter-cell interference by integrating these strategies into Coordinated Multi-Point (CoMP) frameworks.
Summary updated