Abstract
This paper investigates a wireless-powered relaying multiple-input-multiple-output (MIMO) communication system involving four nodes: an information source, an energy source, a relay, and a destination. By employing Orthogonal Frequency Division Multiplexing (OFDM) and a full-duplexing interface at the relay, the system harvests energy from both a dedicated energy source and its own loop-back transmissions. To address the resulting large-scale nonconvex optimization problem for power allocation, the authors propose a high-efficiency path-following algorithm. This algorithm achieves a computational speedup of over $10^5$ compared to standard solvers, enabling practical online implementation for high-spectral-efficiency communication.
Introduction: The Challenge of Powering the IoT
As the Internet of Things (IoT) and 5G networks expand toward trillions of sensors, traditional battery-based power becomes unsustainable. Wireless Power Transfer (WPT) offers a solution, but efficiency remains a hurdle. This research focuses on Full-Duplex (FD) relaying, where a relay simultaneously receives and transmits data. Typically, the relay's own transmission creates "self-interference" that must be canceled. This paper proposes a paradigm shift: treating this interference as a resource for energy harvesting, known as self-energy recycling.
System Architecture and Energy Recycling
The proposed system consists of four multi-antenna nodes. A relay (R) facilitates communication between a source (S) and a destination (D). To sustain its operation, the relay harvests energy from a nearby dedicated energy source (E) and recycles energy from its own output signal.
The communication protocol is divided into two main phases within a time block $T$. In the first half ($T/2$), the source transmits information to the relay. In the second half, the relay forwards this information to the destination while simultaneously harvesting energy from the dedicated source and its own loop-back signal.
Optimization via Path-Following Algorithm
Optimizing power allocation across thousands of OFDM subcarriers and multiple antennas is a complex nonconvex problem. Standard "black-box" solvers (like CVX) are too slow for real-time wireless environments, often taking over an hour for a single calculation.
The authors developed a path-following algorithm that:
- Derives Closed-Form Solutions: By replacing nonconvex functions with specific concave/convex approximations, the algorithm finds analytical solutions at each iteration.
- Achieves Linear Complexity: The computational load grows linearly with the number of subcarriers ($O(KN)$), making it exceptionally fast.
- Enables Real-Time Processing: The algorithm reduces computation time from roughly 75 minutes to just 0.046 seconds—a $10^5$ speedup.
Key Results and Performance
Numerical simulations using 3GPP IMETRA MIMO channel models and up to 4,096 subcarriers demonstrate the system's effectiveness:
- Spectral Efficiency: The optimized power allocation significantly outperforms baseline "Equal Power Allocation" strategies.
- Recycling Gains: Self-energy recycling provides a tangible boost to spectral efficiency, proving that "interference" can indeed be an asset.
- Robustness: The algorithm remains effective even with imperfect channel state information (CSI), which is common in real-world deployments.
Limitations and Future Directions
While promising, the system currently assumes a dedicated energy source is physically close to the relay, which may not always be feasible. Additionally, the hardware required for simultaneous transmission, reception, and energy harvesting is more complex than standard half-duplex components. Future research could extend this framework to multi-user scenarios or multi-relay networks.
Conclusion
By combining self-energy recycling with a hyper-efficient optimization algorithm, this work provides a viable path for self-sustaining, high-performance relaying in future wireless networks. The massive reduction in computational complexity transforms a theoretical model into a practical solution for online power management in complex MIMO-OFDM systems.
Summary updated