Toward a Connected Sky: Performance of Beamforming With Down-Tilted Antennas for Ground and UAV User Co-Existence

Ramy Amer, Walid Saad, Nicola Marchetti

Abstract

As Unmanned Aerial Vehicles (UAVs) become increasingly integrated into cellular networks, the challenge of co-existence with traditional ground users (GUs) intensifies. This paper presents a rigorous analytical framework using stochastic geometry to evaluate the performance of massive MIMO with conjugate beamforming (CB) in such hybrid networks. We demonstrate that antenna down-tilt, typically optimized for ground coverage, can be leveraged to benefit high-altitude aerial users (AUs) by mitigating inter-cell interference. Our findings show that massive MIMO significantly enhances AU connectivity, making cellular-connected UAVs a viable reality without infrastructure overhauls.

Problem Statement: The Challenge of Co-existence

Current cellular infrastructure is inherently "ground-centric." Base station (BS) antennas are typically tilted downwards to focus energy on ground-level users and minimize interference between adjacent cells. However, the rise of UAVs introduces "Aerial Users" (AUs) that operate at high altitudes, often positioned behind or above the main antenna beams.

AUs face a unique interference environment: their clear Line-of-Sight (LoS) to multiple base stations makes them highly susceptible to massive interference from neighboring cells. The primary challenge addressed in this research is the co-existence of AUs and GUs. Specifically, how can we utilize massive MIMO beamforming to serve both user types simultaneously without compromising the spectral efficiency (SE) or connectivity of the existing ground network?

Methodology

The study employs a novel mathematical model based on stochastic geometry to characterize network performance. Key components include:

  • Spatial Modeling: Base stations are modeled as a homogeneous Poisson Point Process (PPP).
  • Channel Modeling: The model incorporates Rayleigh fading for GUs (typical for NLoS urban environments) and Nakagami-$m$ fading for AUs to account for varying LoS/NLoS conditions.
  • Precoding: The focus is on Conjugate Beamforming (CB). While Zero-Forcing (ZF) beamforming is common, CB is computationally simpler and highly effective in massive MIMO regimes where the number of antennas ($M$) greatly exceeds the number of users ($K$).
  • Analytical Metrics: The primary metric derived is the Successful Content Delivery Probability (SCDP), representing the likelihood of achieving a Signal-to-Interference Ratio (SIR) above a target threshold.

Validation and Simulation

This theoretical study validates its analytical derivations through extensive Monte Carlo simulations. The analytical probability density functions (PDFs) for interference channel gains were shown to be highly accurate, particularly as the number of antennas increases, providing a robust basis for the subsequent performance analysis.

Key Insights: The Role of Altitude and Interference

A critical finding is the existence of an altitude threshold. The impact of antenna tilt depends on whether the AU is above or below the base station height ($h_{BS}$):

  • Below $h_{BS}$: A direct tradeoff exists; increasing down-tilt favors GUs but degrades AU performance. For an AU at 30m, the SCDP peaks at a relatively low tilt angle ($\approx 22^\circ$).
  • Above $h_{BS}$: A large down-tilt can create a "win-win" scenario. For an AU at 80m, increasing the down-tilt angle beyond $35^\circ$ actually improves the SCDP for both the AU and the GU by suppressing side-lobe interference from neighboring cells.

Furthermore, AUs exhibit higher interference sensitivity compared to GUs. Due to their LoS conditions, adding more users to the spatial multiplexing mix degrades the AU's performance faster than it does for ground-based users.

Content delivery probability as a function of aerial user altitude, comparing different beamforming strategies.

Performance Results

The proposed MIMO CB strategy shows a substantial performance leap for UAVs compared to traditional single-antenna systems. As shown in the altitude analysis, while AU performance naturally declines as they move further from the main beam at higher altitudes, the use of MIMO CB maintains significantly higher SCDP than non-MIMO configurations.

Impact of system parameters: (a) SCDP vs. scheduled users $K$, (b) Spectral efficiency vs. $K$, and (c) SCDP vs. number of antennas $M$.

Key results from the simulations include:

  • Scaling with Antennas: Increasing the number of antennas ($M$) consistently improves SCDP for both user types, validating the "power of massive MIMO" in this context.
  • User Multiplexing Trade-offs: While increasing the number of scheduled users ($K$) improves the overall system spectral efficiency, it leads to a decrease in individual SCDP, with AUs being more sensitive to this degradation.
  • Spectral Efficiency Gains: Spatially multiplexing an AU with multiple GUs significantly boosts total system throughput compared to serving AUs in isolation.

Conclusion and Future Directions

This research provides a theoretical foundation for integrating UAVs into existing cellular networks. By optimizing antenna tilt and leveraging massive MIMO, operators can support high-flying aerial users alongside ground-based traffic.

Future work should address practical constraints such as imperfect Channel State Information (CSI), which is challenging to maintain for fast-moving UAVs. Additionally, exploring spatial correlation between antenna elements and scenarios with high UAV densities will be crucial for real-world deployment. The potential integration of blockchain for secure content delivery also remains an intriguing avenue for future investigation.

Summary updated

Abstract

Providing connectivity to aerial users (AUs) such as cellular-connected unmanned aerial vehicles (UAVs) is a key challenge for tomorrow's cellular systems. In this letter, the use of conjugate beamforming (CB) for simultaneous content delivery to an AU co-existing with multiple ground users (GUs) is investigated. In particular, a content delivery network of uniformly distributed massive multiple-input multiple-output (MIMO)-enabled ground base stations (BSs) serving both aerial and ground users through spatial multiplexing is considered. For this model, the successful content delivery probability (SCDP) is derived as a function of the system parameters. The effects of various system parameters such as antenna down-tilt angle, AU's altitude, number of scheduled users, and number of antennas on the achievable performance are then investigated. Results reveal that, whenever the AU's altitude is below the BS height, the antennas' down-tilt angles yield an inherent tradeoff between the performance of the AU and the GUs. However, if the AU's altitude exceeds the BS height, down-tilting the BS antennas with a considerably large angle improves the performance of both the AU and the GUs.
BibTeX
@article{amer2019toward,
  author = {Ramy Amer and Walid Saad and Nicola Marchetti},
  title = {{Toward a Connected Sky: Performance of Beamforming With Down-Tilted Antennas for Ground and UAV User Co-Existence}},
  journal = {IEEE Communications Letters},
  year = {2019},
  url = {https://openreview.net/forum?id=xoEudxUf-T9},
}
Figure