PaperlayerView exact revision on arXiv

Source-linked AI summary

Tethered UAVs for Dense Urban Connectivity

German Svistunov, Azim Akhtarshenas

arXiv:2609.10414v1eess.SYcs.SY

TL;DR

The paper examines whether tethered-UAV base stations can provide effective 5G connectivity in dense urban settings where terrestrial networks face coverage limitations. It compares network-wide TUAV-based NTN and terrestrial 5G using system-level evaluation across TUAV altitudes. At 200 m, the NTN scenario has higher median and 5th-percentile throughput than the terrestrial baseline, while higher altitudes can increase interference.

  • Problem

    Terrestrial networks face coverage and capacity demands, and aerial base stations may mitigate coverage limitations caused by shadowing in urban environments.

    Source: §I · p. 1

  • Method

    The study compares large-scale TUAV-based NTN and conventional TN deployments in a dense urban multi-cell scenario, using a low-altitude LoS probability model for dense urban conditions.

    Source: B. Motivation and contributions · p. 2, 3) Channel model · p. 3

  • Results

    At hNTN = 200 m, median UE throughput is 140 Mbps versus 100 Mbps for TN, and 5th-percentile UE throughput is 30 Mbps versus 20 Mbps for TN.

    Source: §III · p. 4

  • Takeaways & Limitations

    An intermediate TUAV altitude offers the best trade-off between improved coverage and interference in the evaluated dense urban scenario.

    Source: §IV · p. 5

Abstract

from arXiv · show

This paper evaluates the downlink performance of 5G non-terrestrial networks (NTNs) realized via tethered unmanned aerial vehicle (TUAV)-mounted base stations, and compares it against conventional 5G terrestrial networks (TNs) in a realistic dense urban scenario. Unlike battery-limited UAVs, TUAVs are connected to ground stations via lightweight cables, enabling stable positioning and near-line-of-sight links to users without the endurance constraints of untethered platforms. Using a 3GPP-compliant multi-cell hexagonal layout with 19 sites and 57 sectors, we model TUAV altitudes ranging from 100 to 1000\,m and evaluate the average effective signal-to-interference-plus-noise ratio (SINR) and user throughput via system-level simulations. Results show that TUAV-based NTN deployments can significantly outperform terrestrial 5G in per-user throughput, with the largest gains observed for cell-edge and low-SINR users, provided the TUAV altitude is properly chosen to balance improved line-of-sight probability against increased propagation loss and interference at higher altitudes.

I. INTRODUCTION

The paper motivates tethered UAV base stations as a way to bring flexible aerial connectivity to dense urban networks and evaluates their performance against terrestrial 5G. It frames the study as a network-wide comparison that examines how uniform TUAV altitude affects performance.

I. INTRODUCTION

  • The paper presents a comprehensive performance evaluation comparing fifth-generation 5G non-terrestrial networks using tethered-UAV aerial base stations with conventional 5G terrestrial networks in a realistic dense urban environment.Source: §I · p. 1

A. Related Works

  • Prior work examines TUAV relay coverage, maritime communications, heterogeneous-network sum rate, placement, trajectory control, intermittent tethering, and hotspot offloading,,,,,,,.Source: A. Related Works · pp. 1–2

B. Motivation and contributions

  • The framework extends the 3GPP channel model with a geometry-based LoS probability model for low-altitude aerial platforms in dense urban clutter.Source: B. Motivation and contributions · p. 2
  • The paper evaluates downlink SINR and UE throughput for TUAV-based NTN and conventional 5G TN in a standardized dense urban multi-cell scenario.Source: B. Motivation and contributions · p. 2
  • The study varies network-wide TUAV altitude from 100 m to 1000 m to examine the trade-off between improved LoS probability, propagation loss, and inter-cell interference.Source: B. Motivation and contributions · p. 2
  • The authors describe this as the first direct comparison of TUAV-based NTN and conventional TN in a standardized, large-scale dense urban multi-cell scenario.Source: B. Motivation and contributions · p. 2

II. SYSTEM MODEL

The system model considers downlink performance for terrestrial and non-terrestrial cellular deployments in an urban macro scenario. It uses a hexagonal network layout and summarizes the simulation parameters.

II. SYSTEM MODEL

  • The model considers downlink performance for terrestrial and non-terrestrial cellular deployments in an urban macro scenario, following the models in,.Source: §II · p. 2
  • Table I summarizes the simulation parameters used in the experiments.Source: Table I · p. 2
  • Figure 1 depicts the network deployment over a hexagonal grid.Source: Figure 1 · p. 2

A. Deployment scenarios

The deployment scenarios use a shared hexagonal multi-cell layout for terrestrial and TUAV-based networks. Outdoor single-antenna UEs are uniformly distributed across the cells, and both scenarios assume full frequency reuse.

1) Network deployment:

2) Frequency bands:

  • Full frequency reuse is assumed in both terrestrial and non-terrestrial scenarios.Source: 2) Frequency bands · p. 2
  • Outdoor single-antenna UEs are uniformly distributed across all cells and each UE associates with one serving base station.The assumption is intended to support evaluation under balanced load conditions.Source: 3) User distribution · p. 3

3) User distribution:

The NTN adjustments orient TUAV antennas toward their target service areas and use a shared conventional UPA configuration for terrestrial and aerial scenarios. The section also introduces a per-UE, per-cell large-scale channel gain.

B. NTN scenario adjustments

1) TUAV-mounted antenna tilt:

  • The mechanical tilt angle is selected so the antenna boresight intersects the center of its target service area, and it depends on TUAV altitude.Source: 1) TUAV-mounted antenna tilt · p. 3

2) UPA antenna configuration:

3) Channel model:

The channel model combines large-scale gains, attenuation, and Rician small-scale fading. It adapts the NTN line-of-sight probability calculation to lower-altitude TUAVs and applies dense-urban parameters.

3) Channel model:

  • Large-scale channel gain combines path-loss, shadowing, and element-antenna gains, using 3GPP statistical channel models for TN and NTN.The model also follows ITU-R recommendations for gaseous, rain, cloud, and scintillation attenuation.Source: 3) Channel model · p. 3
  • Small-scale channel fading is modeled as Rician, with a Rayleigh-faded NLoS component and a plane-wave-approximation LoS component.K denotes the Rician factor.Source: 3) Channel model · p. 3
  • The NTN LoS model uses Sabour et al.’s method to address lower-altitude geometric coverage and urban-clutter effects, with weighted probabilities calculated using dense-urban parameters.Street and crossroad regions are distinguished; the building-size and street-width parameters are defined from the urban scenario characteristics.Source: 3) Channel model · p. 3

C. Cell association, SINR and rate computation

UEs associate with the cell providing the highest RSRP; the model then computes per-layer, per-PRB SINR and effective SINR before deriving downlink rate.

C. Cell association, SINR and rate computation

  • Each UE is associated with the cell providing the highest reference signal received power, computed from average reference-signal power and channel effects.Transmit power and antenna gains are also accounted for.Source: C. Cell association, SINR and rate… · p. 3
  • Per-layer, per-PRB downlink SINR accounts for serving-cell transmission and interference, with transmit power uniformly distributed across allocated PRBs and beams.The SINR calculation also includes noise power.Source: C. Cell association, SINR and rate… · p. 3
  • A mutual-information-based mapping converts per-PRB SINRs into effective SINR, and achievable downlink rate is computed under round-robin scheduling and full-buffer traffic.Effective SINR represents UE quality over allocated frequency resources.Source: C. Cell association, SINR and rate… · p. 4

N PRB

The rate expression uses the serving cell’s available PRBs, PRB bandwidth, and the number of UEs scheduled on each MIMO layer.

N PRB

  • The serving cell’s total number of available PRBs is denoted by N_PRB.Source: N PRB · p. 4
  • B_PRB denotes the bandwidth of each PRB.Source: N PRB · p. 4
  • N_UE for a serving-cell MIMO layer denotes the number of UEs scheduled on that layer.Source: N PRB · p. 4

III. SIMULATION RESULTS

At 200 m, TUAV-based NTN achieves the strongest reported throughput, while higher altitude brings a trade-off between improved line of sight and increased interference and propagation loss.

III. SIMULATION RESULTS

  • The evaluation uses Giulia, a system-level simulator calibrated to 3GPP channel and deployment models, to compare the proposed TUAV-based 5G NTN deployment with conventional terrestrial 5G cellular networks.Source: §III · p. 4
  • At hNTN = 200 m, median UE throughput is 140 Mbps versus 100 Mbps for TN, and 5th-percentile throughput is 30 Mbps versus 20 Mbps.Most NTN scenarios also outperform TN in the low-SINR region.Source: §III · p. 4
  • Both useful and interference power are significantly higher in NTN scenarios, primarily because of increased LoS probability.The 100 m NTN case differs from higher altitudes because the minimum achievable elevation angle increases with altitude.Source: §III · p. 4, Figure 3 · p. 4
  • Raising altitude improves LoS probability, especially near cell edges, but also increases propagation loss and beam footprint, so interference may dominate at higher altitudes.The minimum elevation angle rises from approximately 17° at 100 m to 31° at 200 m, roughly doubling LoS probability from 0.2 to 0.4.Source: §III · pp. 4–5

A. Future Works

Future work will extend the framework to optimize TUAV altitude and placement and evaluate additional user-distribution, mobility, and coordination scenarios.

A. Future Works

  • Future work will explore joint altitude and placement optimization with DRL, non-uniform clustered users and UE mobility, and multi-TUAV coordination to mitigate higher-altitude inter-cell interference.Source: A. Future Works · p. 5

IV. CONCLUSION

In the evaluated dense-urban scenario, TUAV-based 5G NTN can outperform terrestrial 5G in per-user throughput, particularly for users with poor connectivity, when altitude is appropriately chosen.

IV. CONCLUSION

  • TUAV-based 5G NTN can significantly outperform terrestrial 5G in per-user throughput, especially for cell-edge users, with an intermediate altitude balancing coverage and interference.The evaluation considers platform altitudes from 100 m to 1000 m.Source: §IV · p. 5
  • Future work will examine joint TUAV altitude and placement optimization and the effects of non-uniform user distributions and mobility.Source: §IV · p. 5
Loading 2609.10414v1…