Modeling and Analysis of Air-to-Ground Cellular KPIs in a 5G Testbed using Android Smartphones

An{\i}l G\"urses; \.Ismail G\"uven\c{c}; Magreth Mushi; Mihail L. Sichitiu; \"Ozg\"ur \"Ozdemir; Ram Asokan; Rudra Dutta; Simran Singh

arxiv: 2604.04452 · v2 · submitted 2026-04-06 · 📡 eess.SP · cs.PF

Modeling and Analysis of Air-to-Ground Cellular KPIs in a 5G Testbed using Android Smartphones

Simran Singh , An{\i}l G\"urses , \"Ozg\"ur \"Ozdemir , Ram Asokan , Mihail L. Sichitiu , \.Ismail G\"uven\c{c} , Rudra Dutta , Magreth Mushi This is my paper

Pith reviewed 2026-05-10 20:14 UTC · model grok-4.3

classification 📡 eess.SP cs.PF

keywords air-to-groundUAV5GKPIsmodelingmachine learningtestbedAndroid

0 comments

The pith

Real UAV flights with Android phones yield polynomial and ML models for 5G air-to-ground KPI variations with position.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper runs aerial flights at a private 4G/5G testbed to capture how signal strength, throughput, and other KPIs change with a UAV's altitude, distance, elevation, and azimuth relative to the base station. Data collected from two Android devices is compared against free-space path loss predictions that include antenna patterns, after which parameters for polynomial approximations are fitted to the measurements. Light machine learning regressors—random forests, gradient boosting, and neural networks—are then trained to predict KPI values directly from UAV position. These data-driven models are presented as higher-fidelity tools than generic theoretical formulas for designing and simulating cellular-enabled UAV systems.

Core claim

The paper shows that observed variations in 4G and 5G physical-layer KPIs and application throughput can be represented by polynomial curve fits whose parameters are derived from the flight data, and that random forests, gradient boosting regressors, and neural networks can accurately map UAV position relative to the base station onto those KPIs.

What carries the argument

Polynomial curve approximations whose parameters are derived from measured data, together with light machine learning regressors (random forests, gradient boosting, neural networks) that take UAV altitude, distance, elevation, and azimuth as inputs to predict KPI values.

If this is right

UAV mission planners can use the position-to-KPI mappings to select flight paths that maintain reliable command links.
Simulation tools for cellular UAV networks gain accuracy by replacing generic path-loss formulas with the fitted polynomials and trained regressors.
System designers obtain concrete guidance on how elevation and azimuth affect 5G throughput in addition to distance and altitude.
The approach demonstrates that modest amounts of real flight data suffice to build usable predictive models without full ray-tracing.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same measurement campaign could be repeated at increasing altitudes to check whether the polynomial order or ML feature importance remains stable.
Integration of the models into real-time UAV controllers would allow dynamic adjustment of transmit power or handover thresholds based on instantaneous position.
Extending the feature set to include time-of-day or weather variables might further reduce prediction error in operational settings.

Load-bearing premise

The KPI patterns measured at this one testbed with the chosen phones and base station hold for other locations, equipment, and propagation environments.

What would settle it

KPI measurements collected at a second testbed or with different phones and base stations that fall outside the confidence intervals of the derived polynomial fits or ML predictions.

Figures

Figures reproduced from arXiv: 2604.04452 by An{\i}l G\"urses, \.Ismail G\"uven\c{c}, Magreth Mushi, Mihail L. Sichitiu, \"Ozg\"ur \"Ozdemir, Ram Asokan, Rudra Dutta, Simran Singh.

**Figure 1.** Figure 1: AERPAW Lake Wheeler Road Field Labs, where the UAV flights were conducted. The site has rural terrain characteristics. 2. AERPAW 4G/5G Measurement Testbed In this section, we provide details of the AERPAW testbed, configuration of the 4G/5G BS and antenna radiation pattern, and the Android software used to collect radio and throughput measurements. The flights were conducted on AERPAW’s Lake Wheeler Road F… view at source ↗

**Figure 2.** Figure 2: Radiation pattern of the 5G antenna used [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Figure 3: PawPrints architecture and components [17]. (a) Polygonal trajectory around the BS. (b) Two sweeps of the UAV across the BS. (c) Horizontal sawtooth trajectory across the BS [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Figure 4: Flight trajectories used in the measurement [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

**Figure 7.** Figure 7: UAV position and orientation (a), and free () [PITH_FULL_IMAGE:figures/full_fig_p006_7.png] view at source ↗

**Figure 6.** Figure 6: Measured and predicted RSRP during landing [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗

**Figure 8.** Figure 8: UAV position and orientation (a) and free () [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗

**Figure 9.** Figure 9: UAV position and orientation (a) and free () [PITH_FULL_IMAGE:figures/full_fig_p008_9.png] view at source ↗

**Figure 10.** Figure 10: Heatmaps of LTE RSRP (dBm), RSRQ (dB), and cell associations are shown in (a), (b), and (c) [PITH_FULL_IMAGE:figures/full_fig_p009_10.png] view at source ↗

**Figure 11.** Figure 11: The heatmap of 5G SS RSRP is shown for the horizontal sawtooth trajectory at an altitude of 30 m in (a) and at an altitude of 50 m in (b), as measured at Samsung S23 smartphone using Nemo. (a) 5G SS SINR along horizontal sawtooth trajectory at 30 m altitude (b) 5G SS SINR along horizontal sawtooth trajectory at 50 m altitude [PITH_FULL_IMAGE:figures/full_fig_p010_11.png] view at source ↗

**Figure 13.** Figure 13: The heatmap of 5G PDCP throughput is shown for the horizontal sawtooth trajectory at an altitude of 30 m in (a) and at an altitude of 50 m in (b), as measured at Samsung S23 smartphone using Nemo. (a) CQI along horizontal sawtooth trajectory at 30 m altitude (b) CQI along horizontal sawtooth trajectory at 50 m altitude [PITH_FULL_IMAGE:figures/full_fig_p011_13.png] view at source ↗

**Figure 15.** Figure 15: The heatmap of PDSCH channel rank is shown for the horizontal sawtooth trajectory at an altitude of [PITH_FULL_IMAGE:figures/full_fig_p012_15.png] view at source ↗

read the original abstract

The integration of cellular communication with Unmanned Aerial Vehicles (UAVs) extends the range of command and control and payload communications of autonomous UAV applications. Accurate modeling of this air-to-ground wireless environment aids UAV mission planning. Models built on and insights obtained from real-life experiments intricately capture the variations in air-to-ground link quality with UAV position, offering more fidelity for simulations and system design than those that rely on generic theoretical models designed for ground scenarios or ray-tracing simulations. In this work, we conduct aerial flights at the Aerial Experimentation and Research Platform for Advanced Wireless (AERPAW) Lake Wheeler testbed to study the variation in key performance indicators (KPIs) of a private 4G/5G cellular base station (BS) with the UAV's altitude, distance from the BS, elevation, and azimuth relative to the BS. Variations in 4G and 5G physical layer KPIs and application layer throughput are logged and analyzed, using two Android smartphones: a Keysight Nemo device, with enhanced KPI access, through a rooted operating system, and a standard smartphone running a custom application that utilizes open-source Android APIs. The observed signal strength measurements are compared to theoretical predictions from free space path loss models that incorporate the BS antenna radiation patterns. Mathematical model parameters for polynomial curve approximations are derived to fit the observed data. Light machine learning approaches, namely random forests, gradient boosting regressors and neural networks, are used to model KPI behaviour as a function of UAV position relative to the BS. The insights and models generated from real-life experiments in this study can serve as valuable tools in the design, simulation and deployment of cellular communication-based UAV systems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This paper gives useful real-device KPI measurements from one UAV testbed but the position-based polynomial and ML models are fitted to that same data and lack any shown transfer to other sites.

read the letter

This paper supplies concrete KPI measurements from UAV flights at a single 5G testbed using Android phones, but the polynomial and ML models are derived from that same limited data and show no evidence of working elsewhere. They collected 4G and 5G signal and throughput data during controlled flights at the AERPAW Lake Wheeler site. Two Android devices logged the KPIs while the UAV changed altitude, distance, elevation and azimuth relative to a private base station. The measurements are set against free-space path loss predictions that account for the base station antenna patterns. The strength is the real-device data collection. Using both a rooted phone with extra KPI access and a standard phone with open APIs makes the results closer to what a practical UAV system would see. The polynomial curve fits and the random forest, gradient boosting and neural network regressors give position-dependent predictors that match the observed variations at this site better than generic models. The weakness is the lack of broader testing. Everything is from one location and one base station setup, so the fitted coefficients and trained models include the specific propagation effects there. The abstract does not describe separate validation data, uncertainty estimates or checks against other environments. This makes it difficult to judge how useful the models would be for different sites or equipment. Engineers working on UAV mission planning or cellular network simulations for similar rural testbed conditions will find the measurements and the simple predictors useful. The paper does not claim new theory but provides grounded data that simulations often lack. It deserves peer review because the experimental setup is described clearly and the practical goal is well defined, though the referees will likely ask for more on validation and scope.

Referee Report

3 major / 2 minor

Summary. The paper reports KPI measurements (signal strength, throughput) collected during UAV flights at the single AERPAW Lake Wheeler testbed using two Android smartphones over a private 4G/5G BS. It compares observed values to free-space path-loss predictions that incorporate BS antenna patterns, derives polynomial coefficients to fit the altitude/distance/elevation/azimuth dependence, and trains random-forest, gradient-boosting, and neural-network regressors on UAV-position features to model the same KPIs.

Significance. Real smartphone-based air-to-ground KPI traces from a controlled testbed are a useful addition to the literature; if the fitted polynomials and ML mappings were shown to generalize, they could improve fidelity of UAV communication simulations over purely theoretical or ray-tracing models. The explicit comparison against free-space predictions and the use of both enhanced (rooted) and standard Android APIs are concrete strengths.

major comments (3)

[Abstract / ML modeling] Abstract and modeling sections: polynomial coefficients and ML model parameters are obtained by direct fitting to the same flight observations; no cross-validation, hold-out test set, or out-of-sample RMSE is reported, so the claimed “higher-fidelity models” remain untested for predictive accuracy beyond the training flights.
[Abstract / Conclusion] Abstract and conclusion: the utility asserted for “design, simulation and deployment of cellular communication-based UAV systems” rests on an untested assumption that the learned mappings transfer to other BS heights, frequencies, terrain, or multipath environments; no sensitivity analysis or multi-site data is provided.
[KPI collection and analysis] KPI analysis: measurement noise, device-specific biases between the Keysight Nemo and standard smartphone, and flight-to-flight variability are not quantified with error bars or confidence intervals, weakening the comparison to theoretical free-space models.

minor comments (2)

[Experimental setup] Notation for elevation and azimuth angles should be defined explicitly when first introduced to avoid ambiguity with standard spherical-coordinate conventions.
[Figures] Figure captions should state the number of flights, total samples, and any filtering applied so readers can assess statistical robustness.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments and for recognizing the value of real smartphone-based air-to-ground KPI traces from the AERPAW testbed. We address each major comment point by point below.

read point-by-point responses

Referee: [Abstract / ML modeling] Abstract and modeling sections: polynomial coefficients and ML model parameters are obtained by direct fitting to the same flight observations; no cross-validation, hold-out test set, or out-of-sample RMSE is reported, so the claimed “higher-fidelity models” remain untested for predictive accuracy beyond the training flights.

Authors: We agree that the current models are fitted directly to the collected flight data to capture observed position-dependent KPI variations. To strengthen the claims, we will add k-fold cross-validation and report out-of-sample RMSE for the polynomial fits as well as the random forest, gradient boosting, and neural network models in the revised manuscript. revision: yes
Referee: [Abstract / Conclusion] Abstract and conclusion: the utility asserted for “design, simulation and deployment of cellular communication-based UAV systems” rests on an untested assumption that the learned mappings transfer to other BS heights, frequencies, terrain, or multipath environments; no sensitivity analysis or multi-site data is provided.

Authors: Our work is based on a single testbed with a specific BS configuration. We will revise the abstract and conclusion to emphasize that the models and insights are derived from this particular setup and environment. We will add discussion of factors that may affect transferability and identify multi-site validation as future work. revision: yes
Referee: [KPI collection and analysis] KPI analysis: measurement noise, device-specific biases between the Keysight Nemo and standard smartphone, and flight-to-flight variability are not quantified with error bars or confidence intervals, weakening the comparison to theoretical free-space models.

Authors: We will revise the KPI analysis section to quantify variability by adding error bars (standard deviation across repeated flights or measurements) to relevant plots and comparisons with free-space path loss. We will also report observed differences between the two smartphone devices. revision: yes

Circularity Check

0 steps flagged

No significant circularity: empirical fits presented as such

full rationale

The paper collects KPI measurements from UAV flights at a single testbed and states that polynomial parameters 'are derived to fit the observed data' while ML regressors 'are used to model KPI behaviour as a function of UAV position'. This is explicit empirical curve-fitting and regression on the collected dataset, not a first-principles derivation or 'prediction' that reduces to the inputs by construction. The work also compares measurements against independent free-space path-loss models incorporating antenna patterns, providing an external benchmark. No self-citations, uniqueness theorems, or ansatzes imported from prior author work appear in the provided text. The central output is therefore a set of data-driven models whose fidelity claim is evaluated against the same observations and against theoretical baselines, which is self-contained and non-circular.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 0 invented entities

The central claims rest on experimental measurements fitted with polynomials and ML; free parameters are the fit coefficients themselves. The only explicit axiom is the applicability of adjusted free-space path loss as a baseline comparison.

free parameters (2)

polynomial coefficients
Chosen to approximate observed KPI curves versus altitude, distance, elevation and azimuth.
ML model parameters
Trained on the collected flight data to predict KPIs from position.

axioms (1)

domain assumption Free space path loss model adjusted for base-station antenna patterns provides a valid baseline for air-to-ground signal strength
Invoked when comparing measurements to theoretical predictions.

pith-pipeline@v0.9.0 · 5660 in / 1334 out tokens · 71703 ms · 2026-05-10T20:14:39.865340+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages

[1]

Cost-effective measurement setup for an- alyzing signal coverage in 4g/5g mobile networks,

L. Polak, M. Baranek, J. Kufa, R. Sotner, and J. Dluha, “Cost-effective measurement setup for an- alyzing signal coverage in 4g/5g mobile networks,” in 2025 35th International Conference Radioelek- tronika (RADIOELEKTRONIKA), pp. 1–6, 2025

work page 2025
[2]

How Drone, 5G, And AI Technologies Are Enabling First Responders,

Forbes, “How Drone, 5G, And AI Technologies Are Enabling First Responders,” 2025

work page 2025
[3]

UA V Command and Control, Navigation and Surveillance: A Review of Potential 5G and Satellite Systems,

N. Hosseini, H. Jamal, J. Haque, T. Magesacher, and D. W. Matolak, “UA V Command and Control, Navigation and Surveillance: A Review of Potential 5G and Satellite Systems,” in 2019 IEEE Aerospace Conference, pp. 1–10, 2019

work page 2019
[4]

The analysis of key performance indicators (KPI) in 4G/LTE networks,

F. Krasniqi, L. Gavrilovska, and A. Maraj, “The analysis of key performance indicators (KPI) in 4G/LTE networks,” in International Conference on Future Access Enablers of Ubiquitous and Intelli- gent Infrastructures, pp. 285–296, Springer, 2019

work page 2019
[5]

Practical 5G KPI Measurement Results on a Non- Standalone Architecture.,

G. Soós, D. Ficzere, P. Varga, and Z. Szalay, “Practical 5G KPI Measurement Results on a Non- Standalone Architecture.,” in Noms, vol. 2020, pp. 1–5, 2020

work page 2020
[6]

Experimentation and 5g kpi measurements in the 5genesis platforms,

G. Xylouris, M. Christopoulou, H. Koumaras, M.- 7 Table 3: Comparison of KPI values measured during the horizontal sawtooth flights at the two altitudes of 30 m and 50 m. m30 denotes KPI values measured at an altitude of 30 m, m50 denotes KPI values measured at an altitude of 50 m. KPI m30(t) − m50(t) (mean differ- ence) σm30−m50 (standard deviation of di...

work page 2021
[7]

A comprehen- sive overview on 5g-and-beyond networks with uavs: From communications to sensing and intelligence,

Q. Wu, J. Xu, Y. Zeng, D. W. K. Ng, N. Al-Dhahir, R. Schober, and A. L. Swindlehurst, “A comprehen- sive overview on 5g-and-beyond networks with uavs: From communications to sensing and intelligence,” IEEE Journal on Selected Areas in Communica- tions, vol. 39, no. 10, pp. 2912–2945, 2021

work page 2021
[8]

A survey of air- to-ground propagation channel modeling for un- manned aerial vehicles,

W. Khawaja, I. Guvenc, D. W. Matolak, U.-C. Fiebig, and N. Schneckenburger, “A survey of air- to-ground propagation channel modeling for un- manned aerial vehicles,” IEEE Communications Surveys & Tutorials, vol. 21, no. 3, pp. 2361–2391, 2019

work page 2019
[9]

Mobile network performance and tech- nical feasibility of LTE-powered unmanned aerial vehicle,

M. A. Zulkifley, M. Behjati, R. Nordin, and M. S. Zakaria, “Mobile network performance and tech- nical feasibility of LTE-powered unmanned aerial vehicle,” Sensors, vol. 21, no. 8, p. 2848, 2021

work page 2021
[10]

Per- formance of 5G terrestrial network deployments for serving UA V communications,

Z. Huang, J. Rodríguez-Piñeiro, T. Domínguez- Bolaño, X. Yin, J. Lee, and D. Matolak, “Per- formance of 5G terrestrial network deployments for serving UA V communications,” in 2020 14th European Conference on Antennas and Propagation (EuCAP), pp. 1–5, 2020

work page 2020
[11]

Reliable aerial mobile communications with rsrp & rsrq prediction models for the internet of drones: A machine learning approach,

M. Behjati, M. A. Zulkifley, H. A. Alobaidy, R. Nordin, and N. F. Abdullah, “Reliable aerial mobile communications with rsrp & rsrq prediction models for the internet of drones: A machine learning approach,” Sensors, vol. 22, no. 15, p. 5522, 2022

work page 2022
[12]

Air-to-Ground Channel Modeling for UA Vs in Rural Areas,

A. Gürses and M. L. Sichitiu, “Air-to-Ground Channel Modeling for UA Vs in Rural Areas,” in 2024 IEEE 100th Vehicular Technology Conference (VTC2024-Fall), pp. 1–6, 2024

work page 2024
[13]

Fixed-wing uav based air-to-ground channel measurement and modeling at 2.7 ghz in rural environment,

Y. Lyu, W. Wang, and P. Chen, “Fixed-wing uav based air-to-ground channel measurement and modeling at 2.7 ghz in rural environment,” IEEE Transactions on Antennas and Propagation, 2024

work page 2024
[14]

End-to-end performance measurements of drone communications in 5g cellu- lar networks,

A. Festag, S. Udupa, L. Garcia, R. Wellens, M. Hecht, and P. Ulfig, “End-to-end performance measurements of drone communications in 5g cellu- lar networks,” in 2021 IEEE 94th Vehicular Tech- nology Conference (VTC2021-Fall), pp. 1–6, IEEE, 2021

work page 2021
[15]

Aerial Experimentation and Research Platform for Advanced Wireless,

AERPA W, “Aerial Experimentation and Research Platform for Advanced Wireless,” 2024

work page 2024
[16]

A W3232-T0-F Antenna Datasheet,

A. Wireless, “A W3232-T0-F Antenna Datasheet,” 2025

work page 2025
[17]

PawPrints android-based wireless mea- surements tool,

PawPrints, “PawPrints android-based wireless mea- surements tool,” 2025

work page 2025
[18]

Nemo Handy App,

Keysight, “Nemo Handy App,” 2025

work page 2025
[19]

Android Telephony Manager,

Google, “Android Telephony Manager,” 2025. 8 LTE RSRP (dBm) Ericsson cell tower (a) LTE RSRP (dBm) LTE RSRP (dBm) Ericsson cell tower (b) LTE RSRQ (dB) PCI-1 PCI-2 Ericsson cell tower (c) Cell association 0 200 400 600 800 1000 1200 Elapsed time since takeoff (seconds) -115 -110 -105 -100 -95 -90 -85 -80 -75 -70 RSRP (dBm) PCI-1 PCI-2 Connected (d) Variat...

work page 2025
[20]

Lang, Algebra, vol

S. Lang, Algebra, vol. 211 of Graduate Texts in Mathematics. New York, NY, USA: Springer, Revised 3rd ed., 2002. 9 (a) 5G SS RSRP along horizontal sawtooth trajectory at 30 m altitude (b) 5G SS RSRP along horizontal sawtooth trajectory at 50 m altitude Figure 11: The heatmap of 5G SS RSRP is shown for the horizontal sawtooth trajectory at an altitude of 3...

work page 2002
[21]

He is a recipient of Sony Eric- sson’s Distinguished Inventor Award, Ixia Engineering Master Award, and Ixia Technical Excellence Award

Asokan is the inventor/coinventor in some 25 patents and he is a member of Strategic Circle at Asia Open RAN Academy. He is a recipient of Sony Eric- sson’s Distinguished Inventor Award, Ixia Engineering Master Award, and Ixia Technical Excellence Award. At Keysight Technologies, Asokan led wireless test system design and development of the 4G LTE and 5G ...

work page 2001

[1] [1]

Cost-effective measurement setup for an- alyzing signal coverage in 4g/5g mobile networks,

L. Polak, M. Baranek, J. Kufa, R. Sotner, and J. Dluha, “Cost-effective measurement setup for an- alyzing signal coverage in 4g/5g mobile networks,” in 2025 35th International Conference Radioelek- tronika (RADIOELEKTRONIKA), pp. 1–6, 2025

work page 2025

[2] [2]

How Drone, 5G, And AI Technologies Are Enabling First Responders,

Forbes, “How Drone, 5G, And AI Technologies Are Enabling First Responders,” 2025

work page 2025

[3] [3]

UA V Command and Control, Navigation and Surveillance: A Review of Potential 5G and Satellite Systems,

N. Hosseini, H. Jamal, J. Haque, T. Magesacher, and D. W. Matolak, “UA V Command and Control, Navigation and Surveillance: A Review of Potential 5G and Satellite Systems,” in 2019 IEEE Aerospace Conference, pp. 1–10, 2019

work page 2019

[4] [4]

The analysis of key performance indicators (KPI) in 4G/LTE networks,

F. Krasniqi, L. Gavrilovska, and A. Maraj, “The analysis of key performance indicators (KPI) in 4G/LTE networks,” in International Conference on Future Access Enablers of Ubiquitous and Intelli- gent Infrastructures, pp. 285–296, Springer, 2019

work page 2019

[5] [5]

Practical 5G KPI Measurement Results on a Non- Standalone Architecture.,

G. Soós, D. Ficzere, P. Varga, and Z. Szalay, “Practical 5G KPI Measurement Results on a Non- Standalone Architecture.,” in Noms, vol. 2020, pp. 1–5, 2020

work page 2020

[6] [6]

Experimentation and 5g kpi measurements in the 5genesis platforms,

G. Xylouris, M. Christopoulou, H. Koumaras, M.- 7 Table 3: Comparison of KPI values measured during the horizontal sawtooth flights at the two altitudes of 30 m and 50 m. m30 denotes KPI values measured at an altitude of 30 m, m50 denotes KPI values measured at an altitude of 50 m. KPI m30(t) − m50(t) (mean differ- ence) σm30−m50 (standard deviation of di...

work page 2021

[7] [7]

A comprehen- sive overview on 5g-and-beyond networks with uavs: From communications to sensing and intelligence,

Q. Wu, J. Xu, Y. Zeng, D. W. K. Ng, N. Al-Dhahir, R. Schober, and A. L. Swindlehurst, “A comprehen- sive overview on 5g-and-beyond networks with uavs: From communications to sensing and intelligence,” IEEE Journal on Selected Areas in Communica- tions, vol. 39, no. 10, pp. 2912–2945, 2021

work page 2021

[8] [8]

A survey of air- to-ground propagation channel modeling for un- manned aerial vehicles,

W. Khawaja, I. Guvenc, D. W. Matolak, U.-C. Fiebig, and N. Schneckenburger, “A survey of air- to-ground propagation channel modeling for un- manned aerial vehicles,” IEEE Communications Surveys & Tutorials, vol. 21, no. 3, pp. 2361–2391, 2019

work page 2019

[9] [9]

Mobile network performance and tech- nical feasibility of LTE-powered unmanned aerial vehicle,

M. A. Zulkifley, M. Behjati, R. Nordin, and M. S. Zakaria, “Mobile network performance and tech- nical feasibility of LTE-powered unmanned aerial vehicle,” Sensors, vol. 21, no. 8, p. 2848, 2021

work page 2021

[10] [10]

Per- formance of 5G terrestrial network deployments for serving UA V communications,

Z. Huang, J. Rodríguez-Piñeiro, T. Domínguez- Bolaño, X. Yin, J. Lee, and D. Matolak, “Per- formance of 5G terrestrial network deployments for serving UA V communications,” in 2020 14th European Conference on Antennas and Propagation (EuCAP), pp. 1–5, 2020

work page 2020

[11] [11]

Reliable aerial mobile communications with rsrp & rsrq prediction models for the internet of drones: A machine learning approach,

M. Behjati, M. A. Zulkifley, H. A. Alobaidy, R. Nordin, and N. F. Abdullah, “Reliable aerial mobile communications with rsrp & rsrq prediction models for the internet of drones: A machine learning approach,” Sensors, vol. 22, no. 15, p. 5522, 2022

work page 2022

[12] [12]

Air-to-Ground Channel Modeling for UA Vs in Rural Areas,

A. Gürses and M. L. Sichitiu, “Air-to-Ground Channel Modeling for UA Vs in Rural Areas,” in 2024 IEEE 100th Vehicular Technology Conference (VTC2024-Fall), pp. 1–6, 2024

work page 2024

[13] [13]

Fixed-wing uav based air-to-ground channel measurement and modeling at 2.7 ghz in rural environment,

Y. Lyu, W. Wang, and P. Chen, “Fixed-wing uav based air-to-ground channel measurement and modeling at 2.7 ghz in rural environment,” IEEE Transactions on Antennas and Propagation, 2024

work page 2024

[14] [14]

End-to-end performance measurements of drone communications in 5g cellu- lar networks,

A. Festag, S. Udupa, L. Garcia, R. Wellens, M. Hecht, and P. Ulfig, “End-to-end performance measurements of drone communications in 5g cellu- lar networks,” in 2021 IEEE 94th Vehicular Tech- nology Conference (VTC2021-Fall), pp. 1–6, IEEE, 2021

work page 2021

[15] [15]

Aerial Experimentation and Research Platform for Advanced Wireless,

AERPA W, “Aerial Experimentation and Research Platform for Advanced Wireless,” 2024

work page 2024

[16] [16]

A W3232-T0-F Antenna Datasheet,

A. Wireless, “A W3232-T0-F Antenna Datasheet,” 2025

work page 2025

[17] [17]

PawPrints android-based wireless mea- surements tool,

PawPrints, “PawPrints android-based wireless mea- surements tool,” 2025

work page 2025

[18] [18]

Nemo Handy App,

Keysight, “Nemo Handy App,” 2025

work page 2025

[19] [19]

Android Telephony Manager,

Google, “Android Telephony Manager,” 2025. 8 LTE RSRP (dBm) Ericsson cell tower (a) LTE RSRP (dBm) LTE RSRP (dBm) Ericsson cell tower (b) LTE RSRQ (dB) PCI-1 PCI-2 Ericsson cell tower (c) Cell association 0 200 400 600 800 1000 1200 Elapsed time since takeoff (seconds) -115 -110 -105 -100 -95 -90 -85 -80 -75 -70 RSRP (dBm) PCI-1 PCI-2 Connected (d) Variat...

work page 2025

[20] [20]

Lang, Algebra, vol

S. Lang, Algebra, vol. 211 of Graduate Texts in Mathematics. New York, NY, USA: Springer, Revised 3rd ed., 2002. 9 (a) 5G SS RSRP along horizontal sawtooth trajectory at 30 m altitude (b) 5G SS RSRP along horizontal sawtooth trajectory at 50 m altitude Figure 11: The heatmap of 5G SS RSRP is shown for the horizontal sawtooth trajectory at an altitude of 3...

work page 2002

[21] [21]

He is a recipient of Sony Eric- sson’s Distinguished Inventor Award, Ixia Engineering Master Award, and Ixia Technical Excellence Award

Asokan is the inventor/coinventor in some 25 patents and he is a member of Strategic Circle at Asia Open RAN Academy. He is a recipient of Sony Eric- sson’s Distinguished Inventor Award, Ixia Engineering Master Award, and Ixia Technical Excellence Award. At Keysight Technologies, Asokan led wireless test system design and development of the 4G LTE and 5G ...

work page 2001