TL;DR
This paper introduces a sustainable way to power and collect data from battery-free underground sensors using a UAV-enabled "hybrid" wireless energy transfer (WET) system. By optimizing time allocation across charging and communication phases and leveraging "CSI-free" beamforming, the researchers significantly reduce the UAV’s energy consumption while ensuring reliable data transmission from beneath the soil.
What Problem Does This Paper Address?
Wireless Underground Sensor Networks (WUSNs) are vital for smart agriculture and infrastructure monitoring, but they face a major hurdle: power. Replacing batteries in buried devices is impractical, and while wireless energy transfer (WET) from above-ground sources is a solution, it struggles with two issues:
- Extreme Attenuation: Soil absorbs radio signals much more aggressively than air.
- CSI Acquisition Costs: Most efficient wireless charging requires "Channel State Information" (CSI)—knowing exactly how the signal behaves between the transmitter and receiver. In underground environments, gathering this data is energy-intensive and often technically infeasible.
The authors address these challenges by introducing a mobile UAV that acts as both a power source and a data relay, moving closer to sensors to overcome soil path loss.
Methods
The authors propose a four-phase system model:
- UAV Charging: A terrestrial Hybrid Access Point (HAP) wirelessly charges the UAV.
- UD Charging (WET): The UAV flies to the field. In a "hybrid" approach, both the HAP and the UAV transmit energy to the Underground Devices (UDs).
- Data Collection (WIT): UDs use their harvested energy to transmit data back to the hovering UAV.
- Data Offloading: The UAV returns to the HAP to upload the collected data.
Key Technical Innovations: * Hybrid WET: Unlike previous models that use only a fixed station or only a UAV, this system leverages both simultaneously to maximize energy harvest. * CSI-Free Beamforming: To avoid the "CSI tax," the authors evaluate several multi-antenna schemes that don't require channel knowledge, most notably Rotary Antenna Beamforming (RAB), where the UAV's antenna array mechanically rotates to sweep energy across the ground. * Time Allocation Optimization: The authors formulated a convex optimization problem to find the perfect balance of time for each phase (charging vs. transmitting) to minimize the UAV's total energy consumption.
Datasets
The paper does not use a traditional machine learning dataset. Instead, it evaluates the system through simulations of a realistic farming scenario.
- Soil Model: They used the Mineralogy-Based Soil Dielectric Model (MBSDM) to simulate signal attenuation based on real-world variables like Volumetric Water Content (VWC), clay percentage (38%), and burial depth (0.4m).
- Parameters: Simulations involved 64 UDs in a 5m radius circle, with the HAP located 600m away.
Key Insights and Ideas
- The Power of Rotation: Rotary Antenna Beamforming (RAB) emerged as a winner for UAV-based charging. By mechanically rotating the antenna array, the UAV can achieve quasi-omnidirectional coverage with high gain without needing to know exactly where each sensor is or how the channel is behaving.
- Hybrid is Better: Charging UDs from both the HAP (while the UAV is traveling) and the UAV (while it hovers) is significantly more efficient than using either source alone.
- The "CSI-Free" Advantage: While full-CSI beamforming is theoretically more efficient at directing energy, the overhead of getting that CSI in a harsh underground environment often negates the benefits. CSI-free schemes provide a more practical path to "sustainable" networks.
Key Results
- Energy Savings: Under the optimized time allocation, the hybrid WET approach reduced the UAV's energy consumption by 13.43% compared to traditional power source approaches and 5.38% compared to UAV-only charging.
- Depth Resilience: In scenarios with high burial depth (1m), traditional fixed power sources failed to exceed the energy harvesting threshold, whereas the proposed hybrid UAV system successfully powered the sensors.
- Sustainability: The study proved that a UAV can harvest enough energy from a HAP in just 120 seconds to complete a full monitoring cycle of a 64-node underground network.
Limitations and Future Work
- Linear Energy Harvesting: The study assumes a linear model for how sensors harvest energy. In reality, hardware is non-linear (efficiency changes with input power), which the authors plan to address in future work.
- Fixed Trajectory: The UAV's flight path and speed were kept constant. Optimizing the 3D trajectory and flight speed could further reduce energy consumption.
- Single-Antenna UDs: The underground devices were assumed to have only one antenna; exploring multi-antenna UDs could improve performance but would increase device complexity and cost.
Summary updated