Introduction
Wireless Underground Sensor Networks (WUSNs) are essential for monitoring critical infrastructure, such as pipelines, and for applications in precision agriculture and disaster recovery. However, extracting data from buried sensors is a significant challenge due to the extreme signal attenuation caused by soil, water, and man-made structures like asphalt. While terrestrial gateways are the standard solution, they are often unavailable in remote or post-disaster environments. Non-Terrestrial Networks (NTNs)—including Unmanned Aerial Vehicles (UAVs), High-Altitude Platforms (HAPs), and Low Earth Orbit (LEO) satellites—provide a promising alternative for wide-area connectivity.
Modeling the Underground-to-Air Link
A primary contribution of this research is the development of a comprehensive multi-layer channel model. Unlike previous studies that simplify the environment, this model accounts for the complex path a signal takes from a buried device through various media interfaces.
The researchers utilized a custom Monte Carlo simulator to evaluate the performance of different LoRaWAN modulation schemes—standard LoRa and the newer Long Range Frequency Hopping Spread Spectrum (LR-FHSS)—across these complex paths. The model incorporates refraction losses at each interface (e.g., pipe to soil, soil to asphalt, asphalt to air) and integrates 3GPP-standard NTN models for path loss and urban clutter.
Key Insights: Platform Trade-offs and Modulation
The study highlights that the choice of platform and modulation scheme depends heavily on the deployment scenario. While UAVs offer low path loss for short-range monitoring, satellites provide the necessary scale for massive machine-type communication (mMTC) across vast areas.
The Role of LR-FHSS
For HAP and LEO satellite links, LR-FHSS emerges as the superior modulation. By hopping across numerous narrow channels and employing redundant headers, LR-FHSS significantly improves robustness against interference and collisions in dense networks. In contrast, standard LoRa (particularly at higher Spreading Factors like SF12) suffers from long "time-on-air," increasing the likelihood of packet collisions when thousands of devices attempt to communicate simultaneously.
Optimization for UAVs
Interestingly, for low-altitude UAVs, LoRa SF7 is the optimal choice. The shorter distance allows for a sufficient link budget even with a lower spreading factor, and the reduced time-on-air minimizes collisions, leading to a higher success probability (~0.88 in rural areas) compared to more "robust" but slower spreading factors.
Environmental Impacts and Scalability
The research identifies Volumetric Water Content (VWC) in soil as the single most critical environmental factor. High VWC levels can effectively block satellite links, suggesting that sensor transmission schedules should be weather-aware to conserve battery. Furthermore, the study demonstrates that LR-FHSS scales significantly better than standard LoRa, maintaining a 12% higher packet delivery ratio in scenarios with up to 100,000 devices.
Conclusion
This analysis confirms that NTNs can effectively support subterranean mMTC, provided the right modulation and platform are chosen for the environment. While UAVs are ideal for localized, high-reliability tasks, LR-FHSS-enabled satellites are essential for large-scale, remote monitoring. Future work will address the impact of Doppler shifts from moving satellites and the requirements for downlink communication.
The simulator used in this study is available on GitHub.
Summary updated