Introduction
The rapid expansion of the Internet of Things (IoT) presents a significant challenge in powering billions of battery-limited devices. Simultaneous Wireless Information and Power Transfer (SWIPT) has emerged as a promising solution, allowing devices to harvest energy from the same RF signals used for data transmission. When combined with Non-Orthogonal Multiple Access (NOMA), which enhances spectral efficiency by allowing multiple users to share the same frequency, the potential for high-capacity, self-sustaining networks is significant.
However, both technologies are highly sensitive to distance; signal strength for both decoding and energy harvesting drops sharply as devices move away from a central base station. Furthermore, most theoretical models assume perfect Channel State Information (CSI), which is rarely achievable in practice. This paper addresses these gaps by integrating a Distributed Antenna System (DAS)—where remote radio units (RRUs) are spread throughout a cell—to "bring the antennas to the users."
System Model and Methodology
The research models a system featuring a central IoT controller and multiple RRUs. The focus is on a two-user NOMA cluster consisting of a "strong" user (near the center) and a "weak" user (near an RRU). As shown in Figure 1, each user employs a power splitter to divide the received signal between Information Decoding (ID) and Energy Harvesting (EH).
Key technical contributions include:
- Closed-form Rate Expressions: Mathematical derivations for data rates that explicitly account for channel estimation errors.
- Joint Optimization: The framework optimizes power splitting ratios and power allocation for two distinct goals: Max-Sum-Rate (maximizing total throughput) and Max-Min Fairness (ensuring equitable service for the worst-off user).
- Spatial Advantage: Users are modeled in specific regions to test how the proximity of RRUs compensates for path loss and imperfect CSI.
Key Insights and Results
The study demonstrates that RRUs serve as essential local power sources, particularly for "weak" users who would otherwise receive negligible energy from a distant central controller. A major insight is that strong users can afford to allocate more power to energy harvesting (lower $\alpha$) due to their robust channel conditions, while weak users must prioritize decoding.
Throughput and Fairness Gains
The proposed SWIPT-NOMA in DAS configuration shows substantial performance improvements over traditional setups. As illustrated in Figure 5, the system achieves a sum-rate of approximately 10 Mbps at a 30W power budget. This significantly outperforms standard centralized SWIPT-NOMA, which only reaches approximately 3 Mbps under similar conditions.
Resilience to Uncertainty
The DAS-NOMA combination proves remarkably robust. Even with imperfect CSI, the system often outperforms centralized NOMA with perfect CSI. The physical proximity of RRUs effectively compensates for the lack of precise channel knowledge, maintaining a minimum harvested energy of 10 mW for all users while optimizing data rates.
Conclusion and Future Work
Integrating DAS into SWIPT-enabled NOMA networks effectively overcomes the limitations of distance and imperfect channel knowledge. While this study focuses on a two-user pairing, future research will address the complexities of multi-user clusters in dense IoT environments and the impact of user mobility on maintaining reliable CSI. Additionally, accounting for hardware non-linearities in RF-to-DC conversion remains a critical step for real-world implementation.
Summary updated