When we design cellular networks for ground users, we rely heavily on a hidden ally: physical clutter. Buildings, trees, hills, and vehicles act as natural shields, blocking signals from distant base stations and allowing the network to reuse the same frequencies just a few hundred meters away without causing chaos. This phenomenon is known as signal shadowing.
However, when an Unmanned Aerial Vehicle (UAV) ascends into the sky, it leaves this protective clutter behind. This transition introduces a fundamental challenge in aerial communications: the altitude paradox.
The Altitude Paradox
As a drone flies higher, the probability of establishing a clear, unobstructed Line-of-Sight (LoS) link to the ground increases dramatically. Intuitively, one might think that a perfect line-of-sight is entirely beneficial, as it yields a highly robust signal from the serving base station.
The paradox, as formulated by Rosas in an early analytical framework, is that this clear view is a double-edged sword. While the desired signal path improves, the drone simultaneously establishes unobstructed LoS links to dozens of neighboring, distant base stations.
[ Ground User ] ----x (Building Blocks Signal) x---- [ Distant Base Station ]
[ UAV ] ______________________________________/ (Unobstructed LoS Interference)
\__________________________________________ [ Serving Base Station ]
As Zeng et al. observe, the complete absence of ground-level obstacles strips the network of its natural signal shadowing. Instead of communicating with a single isolated cell, a high-altitude drone is exposed to an entire regional network simultaneously, resulting in severe multi-cell interference.
The Downlink Dilemma: High Power, High Noise
In the downlink direction (base station to user), this lack of geographic shielding severely degrades the Signal-to-Interference-plus-Noise Ratio (SINR). Because cellular networks reuse the same frequency bands across different cells, a drone flying at 60 to 100 meters is bombarded by overlapping transmissions.
Empirical field studies by Gharib et al. demonstrate that flying higher frequently leads to worse communication quality, even when the received signal power remains high. The receiver simply cannot distinguish its intended signal from the massive wall of overlapping background noise generated by distant towers.
This visibility creates several distinct network management problems:
- Sidelobe Exposure: Terrestrial base station antennas are tilted downward to focus their energy on ground users. Drones flying at high altitudes miss these main beams entirely and are instead illuminated by weaker, unpredictable antenna sidelobes, making interference prediction crucial (Galkin et al.).
- Identifier Collisions: Because the drone can see so far into the distance, it often detects multiple distant towers that share the exact same Physical Cell Identifier (PCI). As Fonseca et al. point out, these PCI collisions lead to severe synchronization errors and terminal connection failures.
- Handover Instability: Without clear cell boundaries, drones experience rapid, unstable handovers as they struggle to lock onto a single serving cell amid a sea of equally visible competitors (Gharib et al.).
The Uplink Problem: Illuminating the Ground Network
The line-of-sight interference problem is entirely symmetric. When a drone transmits data back down to the network (uplink), its high-altitude vantage point means its signal radiates across a massive geographic footprint.
Instead of reaching only its serving base station, the drone's uplink transmission spills over into numerous neighboring cells, blinding their receivers and overwhelming the weaker signals of ground-level users (Challita et al.; Kang et al.). This creates a fundamental operational tension. As Shamsoshoara et al. explain, a drone requiring high power to transmit high-throughput aerial data will inherently generate destructive uplink interference for the existing terrestrial network, reducing overall capacity for ground users.
Why Traditional Network Design Fails
Because line-of-sight propagation alters the physics of the environment, standard terrestrial network engineering principles break down in the sky.
The Density Reversal
In conventional cellular planning, if you need more capacity, you build more base stations. Denser base station deployments mean shorter propagation distances and better throughput for ground users.
For aerial users, however, mathematical modeling by Yu et al. reveals that the ideal density of ground base stations actually decreases as a drone's altitude increases. Because higher altitudes eliminate blockages so effectively, adding more base stations merely scales the volume of interfering signals faster than it improves the desired signal strength.
Channel Homogeneity and NOMA Failure
This line-of-sight dominance also disrupts advanced multi-user access schemes like Non-Orthogonal Multiple Access (NOMA). NOMA allows multiple users to share the same frequency simultaneously by separating their signals in the power domain. This separation relies entirely on distinct differences in the users' natural channel gains (e.g., one user has a strong, clear path, while another has a weak, heavily obstructed path).
As Liu et al. demonstrate, because the aerial environment is uniformly dominated by clear line-of-sight links, the natural channel differences between multiple aerial users become minimal. Without these distinct channel variations, the receiver cannot cleanly decode and separate the overlapping signals, causing the mathematical foundation of NOMA to collapse without artificial channel manipulation.
Engineering Solutions for Coexistence
To integrate drones safely into existing 5G and beyond infrastructure, researchers are developing architectures designed to mitigate or actively exploit this high-altitude visibility.
1. Up-Tilted Antennas and Coordination
Since traditional base stations point down, one direct solution is to supplement them with dedicated, up-tilted antennas optimized for the sky. Chowdhury et al. demonstrate that combining up-tilted antenna arrays with enhanced Inter-Cell Interference Coordination (eICIC) allows networks to carve out clean, dedicated spatial channels for drones without disrupting terrestrial traffic.
2. Coordinated Multipoint (CoMP) Transmission
Instead of fighting the fact that a drone sees multiple base stations, Coordinated Multipoint (CoMP) frameworks turn this visibility into an advantage. Sun et al. show that by mathematically coordinating transmissions across neighboring towers, multiple base stations can send data to a vulnerable aerial user simultaneously. This transforms what would have been destructive inter-cell interference into a constructively combined, highly reliable signal.
3. Intelligent Reflecting Surfaces (IRS)
To combat the channel homogeneity that ruins schemes like NOMA, researchers are introducing Intelligent Reflecting Surfaces (IRS) into multi-UAV deployments. By dynamically shifting the phase of reflecting elements on buildings, an IRS can artificially alter the wireless paths. As Mu et al. demonstrate, this allows the network to deliberately engineer the channel differences required for signal separation, while simultaneously creating destructive interference patterns to cancel out unwanted aerial noise.
Key Foundational Work
Coexistence of Terrestrial and Aerial Users in Cellular Networks (2017) — contributed 2 concept elements to this direction.
UAV Communications Based on Non-Orthogonal Multiple Access (2018) — contributed 1 concept elements to this direction.