TL;DR
This paper compares tethered UAVs (T-UAVs) and untethered UAVs (U-UAVs) for boosting cellular coverage in high-traffic "hot-spot" areas. It finds that T-UAVs can outperform U-UAVs, especially in challenging urban environments, provided they have adequate ground station access and tether length.
What Problem Does This Paper Address?
Modern cellular networks are increasingly strained by "hot-spots" – areas with concentrated user demand, often temporary (e.g., events, emergencies). While UAVs offer a flexible solution to augment terrestrial base stations (TBSs) in these areas, untethered UAVs (U-UAVs) face limitations due to battery life and backhaul capacity. Tethered UAVs (T-UAVs) offer a continuous power and data link via a cable from a ground station (GS), potentially overcoming these U-UAV limitations. However, T-UAVs introduce their own constraints: tether length, GS accessibility, and limited mobility. This paper aims to provide a comprehensive performance comparison between these two UAV deployment strategies in hot-spot scenarios, considering practical deployment challenges.
Methods
The authors employ stochastic geometry to model the random distribution of users within a hot-spot region. They derive joint probability density functions (PDFs) for the distances between the TBS, UAV, and users. A user association policy is developed based on maximizing the signal-to-noise ratio (SNR) for users, defining specific regions where users will connect to either the TBS or the UAV. The coverage probability for both terrestrial and aerial links is then analytically derived. A key contribution is the analytical proof that the optimal hovering location for a T-UAV lies on a specific surface of a spherical cone defined by the GS location and tether constraints.
Datasets
Not discussed in the paper. The analysis relies on theoretical modeling and simulations based on defined system parameters.
Key Insights and Ideas
- T-UAVs offer a compelling alternative: In scenarios with sufficient GS accessibility and tether length, T-UAVs can provide superior coverage compared to U-UAVs, particularly in complex urban environments.
- Trade-offs are critical: The optimal location for a U-UAV involves a delicate balance between maximizing the access link (UAV-to-user) and the backhaul link (TBS-to-UAV). This often leads to a location closer to the TBS.
- T-UAV mobility is constrained but manageable: While T-UAVs are restricted by tether length and GS location, their optimal deployment can be found on a specific surface of a spherical cone, significantly reducing the search space.
- Urban environments pose unique challenges: High-rise buildings and dense urban layouts negatively impact coverage for both UAV types, but T-UAVs show greater resilience due to their stable backhaul.
- GS accessibility and tether length are key T-UAV enablers: The performance gains of T-UAVs are directly tied to the availability of accessible GS locations and the flexibility provided by a longer tether.
Key Results
- Analytical derivations for coverage probabilities of T-UAV and U-UAV assisted systems are presented.
- It is proven that the optimal T-UAV location lies on a portion of the spherical cone surface.
- Numerical results demonstrate that T-UAVs outperform U-UAVs when GS accessibility and tether length are sufficient.
- In dense urban environments, T-UAV and U-UAV performance is comparable, but in high-rise urban environments, T-UAVs show a significant advantage.
- The paper quantifies the impact of tether length, U-UAV availability, and building accessibility on overall system coverage. For instance, to outperform an 80% available U-UAV, a T-UAV might require a tether length of over 45m and 10% building accessibility in a dense urban setting.
Limitations and Future Work
- Interference: The current analysis neglects interference between users and between UAVs/TBSs. The authors acknowledge that interference plays a significant role, especially in urban environments, and suggest it as a direction for future work.
- User Distribution: The primary analysis assumes a uniform user distribution. While the impact of Gaussian distribution is briefly explored, a more comprehensive study across various user distributions would be beneficial.
- Mobility: The current work focuses on static optimal deployment. Future research could explore dynamic deployment and handover strategies for both UAV types.
- GS Location Optimization: While GS accessibility is considered, optimizing GS placement itself is not the focus.
- T-UAV Tether Dynamics: The model assumes a perfectly stretched tether up to its maximum length. Real-world dynamics like wind and gravity could influence actual reachability.
Summary updated