Pith. sign in

REVIEW 3 major objections 5 minor 25 references

A Versatile and Programmable UAV Platform for Radio Access Network and End-to-End Cellular Measurements

T0 review · 3 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A drone-borne LTE probe shows rural signal strength rising with altitude while signal quality falls.

desk verdict A genuinely useful UAV cellular measurement platform that reproduces a known altitude-quality tradeoff; the causal story (LoS gain, neighbor interference) is asserted rather than tested, and the open-source artifacts are missing. read the letter →

arxiv 2509.03818 v1 pith:WGLMPAIP submitted 2025-09-04 cs.NI cs.SYeess.SY

Sherwan Jalal Abdullah , Sravan Reddy Chintareddy , Victor S. Frost , Shawn Keshmiri , Morteza Hashemi This is my paper
classification cs.NIcs.SYeess.SY
keywords UAVcellularmeasurementsLTERSRP/RSRQ/SINRruralnetworkcoverageaerialinterferenceend-to-endthroughputopen-sourcemeasurementplatform
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper builds and tests an unmanned aerial vehicle (UAV) platform that measures both radio-access-network signals and end-to-end network performance in rural areas where drive testing is impractical. Field experiments near Clinton Lake, Kansas, show an altitude-dependent tradeoff: received signal power (RSRP, RSSI) improves at higher altitudes due to better line-of-sight, but signal quality (RSRQ, SINR) degrades because elevated positions expose the modem to more interference from neighboring cells. The platform also captures throughput and latency, reporting that most of the tested area had acceptable signal quality, over 90% supported more than 5 Mbps in both directions, and more than 80% of round-trip times stayed below 150 ms. The paper argues this open, programmable approach fills a gap left by crowd-sourced coverage maps and proprietary or SDR-based measurement tools.

What carries the argument

The platform's core is an integrated UAV system: an NVIDIA Jetson Orin onboard computer running ROS, a flight controller supplying GPS/IMU/pressure data, and a Microhard pMLTE commercial LTE modem. Custom logging software, deployed on the Orin, queries the modem over Ethernet at one-second intervals to record RSRP, RSRQ, RSSI, SINR, PCI, Cell ID, LAC, and neighbor-cell measurements, while Nping and iPerf3 provide RTT and bidirectional throughput to a dedicated server. Geospatial 2D/3D maps and statistical CDF/PDF analyses turn these logs into altitude-resolved coverage and quality assessments.

What would settle it

Fly the same platform at constant horizontal distance from a known base station and vary only altitude, while logging serving and neighbor cell RSRP, RSRQ, and SINR; if RSRP no longer rises with altitude or the RSRQ/SINR drop disappears once neighbor-cell received power is accounted for, the paper's altitude-tradeoff explanation is not supported.

Watch

Extended reading notes

Core claim

The central claim is that a UAV equipped with a commercial LTE modem and a custom onboard logger can simultaneously capture Radio Access Network metrics (RSRP, RSRQ, RSSI, SINR, cell IDs, neighbor-cell data) and end-to-end performance metrics (RTT, uplink and downlink throughput) in rural terrain. The experiments show an altitude-dependent tradeoff: higher altitude improves received signal power due to enhanced line-of-sight conditions, while degrading signal quality because of increased interference from neighboring cells. The paper further concludes that, in the area tested, the network delivered acceptable signal quality over roughly 85% of the region, throughput above 5 Mbps in both dire

Load-bearing premise

The altitude-dependent RSRP gain is attributed to improved line-of-sight, but distance to the serving base station was not held constant across altitudes, so the gain could partly or wholly be a distance effect.

Editorial extensions

If this is right

  • Rural carriers can use this platform to build three-dimensional coverage maps without road access, addressing terrain and sparse-population gaps in crowd-sourced data.
  • The measured altitude tradeoff implies that aerial users may need different cell-selection or handover rules than ground users, since stronger signal power can coincide with worse signal quality.
  • The open-source logger offers a lower-cost, customizable alternative to licensed commercial testing tools and to SDR systems that capture only basic signal strength.
  • Coverage planning should weigh both signal-strength metrics and the spatial extent over which a cell actually serves the area, not just per-cell RSRP averages.
  • The framework can quantify both radio-layer coverage and application-level experience (throughput, latency) in one flight, supporting more complete rural network assessment.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the altitude-driven RSRP gain holds after controlling for distance to the serving cell, it would strengthen the case for designing UAV-specific interference mitigation, such as directional antennas or altitude-aware frequency planning, for aerial connectivity.
  • The same measurement method could be extended to predict ground-level coverage in inaccessible rural areas from aerial samples, though the paper only lists that as future work.
  • The neighbor-cell logs suggest handover prospects are generally good at altitude because at least one neighbor cell remains strong, but whether that holds in more obstructed or denser environments remains an open question the paper does not test.
  • The reported throughput and RTT numbers are a snapshot from one rural site; repeating the flights across seasons and different terrain would clarify how vegetation and topography shift the altitude tradeoff.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper presents a UAV-mounted cellular measurement platform combining a Jetson Orin, a Microhard pMLTE commercial modem, and custom data-logging software to collect RAN metrics (RSRP, RSRQ, RSSI, SINR) and end-to-end metrics (throughput, RTT) in a rural test site near Clinton Lake, Kansas. Experimental results show that RSRP improves with altitude, while RSRQ and SINR degrade; the authors attribute this to improved line-of-sight to the serving cell and increased neighboring-cell interference at altitude. They report that about 85% of the area has acceptable RSRQ, more than 90% of measurements exceed 5 Mbps in both directions, and more than 80% of RTT samples are below 150 ms.

Significance. If the platform and findings are validated, this is a useful open-source contribution: it addresses a real gap in rural cellular measurement, combines radio-layer and application-layer data at 1 s resolution, and avoids proprietary drive-test tools. The paper explicitly offers reproducibility advantages over commercial solutions and prior SDR-based approaches. The altitude-dependent tradeoff and the observation that strong cell-level signal metrics do not guarantee spatial coverage are potentially interesting for UAV communications and network planning. However, the central physical interpretation is currently under-supported, so the significance is conditional on additional analysis.

major comments (3)
  1. [Section III-A, Fig. 4] The claim that RSRP improves with altitude 'primarily' due to enhanced LoS is not established because the experiment does not control for distance to the serving cell or terrain. If the UAV is closer to the cell or over open ground at higher altitudes, the RSRP gain could be a propagation-distance effect rather than LoS. The authors should stratify RSRP by horizontal distance to the serving cell (or include path-loss analysis and terrain profile) before attributing the trend to LoS.
  2. [Section III-A, Fig. 8] The degradation of RSRQ/SINR with altitude is attributed to increased neighboring-cell interference, but no altitude-resolved neighbor-cell RSSI or interference measurement is presented. Fig. 8 shows only aggregate neighbor statistics across the whole flight. The authors should show that neighbor-cell received power increases with altitude and that RSRQ/SINR degrades when neighbor RSSI is higher, or explicitly soften the causal claim.
  3. [Section III-A/III-B] No statistical detail is provided for the altitude trends: number of samples per altitude bin, variance, or confidence intervals are missing. Since Fig. 4 is the key evidence for the tradeoff, the authors should report sample sizes and dispersion so readers can judge whether the trend is robust.
minor comments (5)
  1. [Title and throughout] 'UA V' appears with spurious spaces repeatedly; should be 'UAV'.
  2. [Section III-B, Fig. 5] The threshold is attributed to 'Teltonica Networks'; the correct spelling is Teltonika. Also, the CDF is over measurement samples, not strictly 'area'; please rephrase as '15% of measurement locations' or provide an area-weighted analysis.
  3. [Section III-C] The text says 'more than 90% of the coverage region supports throughput...' but the CDF is over samples/locations. Rephrase to avoid implying spatial interpolation.
  4. [Section II-A and II-B] The paper would benefit from a table listing all collected metrics and their units, and from stating the number of flights, flight duration, and weather conditions to support reproducibility.
  5. [Section III-A, Fig. 3] The caption says 'altitude above sea level' but the text discusses altitude in terms of flight altitude above ground; clarify which reference is used.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper reports an empirical measurement platform and observational results; no fitted parameters are relabeled as predictions and no load-bearing claim reduces to its own inputs.

full rationale

The paper is an instrumentation-and-measurement study. Its central claims are that the UAV platform can collect RAN and end-to-end metrics simultaneously and that field data show an altitude-dependent tradeoff (RSRP improving with altitude, RSRQ/SINR degrading). There is no derivation chain, no fitting procedure, and no predictive model. The altitude-tradeoff statement is presented as an interpretation of plotted measurements (Section III-A, Figures 3-4), not as a quantity derived from equations. The RSRQ poor-quality threshold of -19 dB is taken from an external source, Teltonika Networks [24], and is not fitted to the data. No parameter is fit to one subset and then used to 'predict' a closely related quantity. The self-citations that appear ([5], [6]) concern spectrum sensing and UTM systems and are not load-bearing for the cellular measurement claims. The only potential concern is that the interpretation attributes RSRP gains to improved line-of-sight and RSRQ/SINR loss to neighboring-cell interference without controlling for distance to the serving cell; however, that is an empirical-identification/correctness issue about causal attribution, not circularity. Under the definition used here, circularity requires a prediction or derivation that is equivalent by construction to its inputs, and no such step exists in the paper.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new theoretical entities or fitted physical parameters. The only hand-chosen threshold is the RSRQ quality cutoff, which is adopted from an external vendor and affects the reported coverage percentages. The core assumptions are about the fidelity of off-the-shelf modem measurements and the interpretation of altitude trends as LoS/interference effects.

free parameters (1)
  • RSRQ poor-quality threshold = -19 dB
    Adopted from Teltonika Networks [24] for classifying poor signal quality; it directly determines the reported 15% poor-quality fraction.
assumptions (3)
  • domain assumption The pMLTE modem's reported RSRP, RSRQ, RSSI, and SINR accurately represent the radio environment at the UAV's position.
    The entire analysis rests on the fidelity of COTS modem measurements, which can be affected by antenna orientation, body shadowing, and receiver algorithms (Section II-A).
  • domain assumption The altitude-dependent RSRP improvement is attributable to enhanced line-of-sight rather than changing distance to the serving cell.
    Section III-A interprets the observed improvement as LoS-driven without controlling for varying distances to base stations across the flight trajectory.
  • domain assumption The area-based coverage statistics (e.g., 85% acceptable) are generalizable from the specific flight path.
    The paper computes 'area' fractions from a limited set of flight trajectories without stating flight coverage per altitude.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Versatile and Programmable UAV Platform for Radio Access Network and End-to-End Cellular Measurements." pith.science (2026). https://pith.science/paper/WGLMPAIP

@misc{pith2026250903818,
  author       = {Pith},
  title        = {Pith review of: A Versatile and Programmable UAV Platform for Radio Access Network and End-to-End Cellular Measurements},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WGLMPAIP}},
  note         = {Machine review of arXiv:2509.03818}
}
read the original abstract

In this work, we develop a measurement platform to capture mobile network performance metrics including coverage and quality of service in regions where conventional coverage testing approaches are frequently time-intensive, labor-demanding, and occasionally hazardous. Traditionally, crowd-sourcing methods are used to collect cellular network performance metrics. However, these approaches are inadequate in rural areas due to low-density population, and difficult terrain. The platform described here is a UAV-based and is designed to investigate the mobile network performance through aerial operations and gather Radio Access Network (RAN) signal alongside end-to-end network performance metrics. Our platform gathers metrics through the integration of an onboard computation unit and commercial off-the-shelf cellular modem. The gathered data are subsequently analyzed and displayed using geospatial mapping utilities and statistical techniques to deliver key observations on cellular network performance. Experimental results showed that the received signal power improves at higher altitudes due to enhanced line-of-sight (LoS) conditions as expected. However, the signal quality degrades as a result of increased interference from neighboring cells. The analysis reveals that for most of the geographic area covered in the initial experiments the system maintained acceptable signal quality, with adequate throughput performance for both uplink and downlink communications, while maintaining satisfactory round-trip time characteristics. Notably, the experiment showed that a strong radio signal metric for a given cell does not necessarily translate to consistent spatial coverage across the tested region.

Figures

Figures reproduced from arXiv: 2509.03818 by the authors.

Figure 1
Figure 1. Overall system model and measurement scenario [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Internal system components of the UAV equipped [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 5
Figure 5. CDF of RSRQ measurement demonstrating signal [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figures from the paper (3 more)
Figure 6
Figure 6. Figure 6: PDF of serving cell coverage dominance, with one cell [PITH_FULL_IMAGE:figures/full_fig_p004_6.png]
Figure 7
Figure 7. Figure 7: Comparative RSSI (dBm) performance showing signal [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]
Figure 8
Figure 8. Figure 8: Statistical analysis of neighbor cells RSSI measure [PITH_FULL_IMAGE:figures/full_fig_p005_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

25 extracted references · 25 canonical work pages

  1. [1]

    Experimental Evaluation of the Performance of UA V-assisted Data Collection for Wake-up Radio- enabled Wireless Networks,

    A. V . Sheshashayee, M. Bordin, P. B. del Prever, D. Villa, H. Cheng, C. Petrioli, T. Melodia, and S. Basagni, “Experimental Evaluation of the Performance of UA V-assisted Data Collection for Wake-up Radio- enabled Wireless Networks,” in 2024 IEEE 99th Vehicular Technology Conference (VTC2024-Spring), 2024, pp. 01–06

    work page 2024

  2. [2]

    Coverage is Not Binary: Quantifying Mobile Broadband Quality in Urban, Rural, and Tribal Contexts,

    V . Adarsh, M. Nekrasov, U. Paul, T. Mangla, A. Gupta, M. Vigil-Hayes, E. Zegura, and E. Belding, “Coverage is Not Binary: Quantifying Mobile Broadband Quality in Urban, Rural, and Tribal Contexts,” in 2021 International Conference on Computer Communications and Networks (ICCCN), 2021, pp. 1–9

    work page 2021

  3. [3]

    CellMapper, “Online,” Available https://www.cellmapper.net

    work page

  4. [4]

    OpenSignal, “Online,” Available https://www.opensignal.com/

    work page

  5. [5]

    Col- laborative wideband spectrum sensing and scheduling for networked UA Vs in UTM systems,

    S. R. Chintareddy, K. Roach, K. Cheung, and M. Hashemi, “Col- laborative wideband spectrum sensing and scheduling for networked UA Vs in UTM systems,” in GLOBECOM 2023-2023 IEEE Global Communications Conference. IEEE, 2023, pp. 3064–3069

    work page 2023

  6. [6]

    Federated learning-based collaborative wideband spectrum sens- ing and scheduling for uavs in utm systems,

    ——, “Federated learning-based collaborative wideband spectrum sens- ing and scheduling for uavs in utm systems,” IEEE Transactions on Machine Learning in Communications and Networking , 2025

    work page 2025

  7. [7]

    Public LTE network measurements with drones in rural environment,

    J. Sae, R. Wir ´en, J. Kauppi, H.-L. Maattanen, J. Torsner, and M. Valkama, “Public LTE network measurements with drones in rural environment,” in 2019 IEEE 89th Vehicular Technology Conference (VTC2019-Spring). IEEE, 2019, pp. 1–5

    work page 2019

  8. [8]

    Pathloss Measurements and Modeling for UA Vs Connected to Cellular Networks,

    R. Amorim, P. Mogensen, T. Sorensen, I. Z. Kovacs, and J. Wigard, “Pathloss Measurements and Modeling for UA Vs Connected to Cellular Networks,” in 2017 IEEE 85th Vehicular Technology Conference (VTC Spring), 2017, pp. 1–6

    work page 2017

Show all 25 references
  1. [9]

    Measured Uplink Interference Caused by Aerial Vehicles in LTE Cellular Networks,

    R. Amorim, H. Nguyen, J. Wigard, I. Z. Kov ´acs, T. B. Sørensen, D. Z. Biro, M. Sørensen, and P. Mogensen, “Measured Uplink Interference Caused by Aerial Vehicles in LTE Cellular Networks,” IEEE Wireless Communications Letters, vol. 7, no. 6, pp. 958–961, 2018

    2018

  2. [10]

    InfoVista, “Online,” Available https://www.infovista.com/products/ tems-suite/mobile-network-testing

  3. [11]

    QualiPoc, “Online,” Available https://www.rohde-schwarz

    Rohde and S. QualiPoc, “Online,” Available https://www.rohde-schwarz. com/us/products/test-and-measurement/network-data-collection/ qualipoc-android 63493-55430.html

  4. [12]

    Nemo, “Online,” Available https://www.keysight.com/zz/en/products/ nemo-wireless-network-solutions.html

    K. Nemo, “Online,” Available https://www.keysight.com/zz/en/products/ nemo-wireless-network-solutions.html

  5. [13]

    Evaluating LTE Coverage and Quality from an Unmanned Aircraft System,

    Nekrasov, Michael and Adarsh, Vivek and Paul, Udit and Showalter, Es- ther and Zegura, Ellen and Vigil-Hayes, Morgan and Belding, Elizabeth, “Evaluating LTE Coverage and Quality from an Unmanned Aircraft System,” in 2019 IEEE 16th International Conference on Mobile Ad Hoc and ...

    2019

  6. [14]

    Mobile Network-Connected Drones: Field Trials, Simu- lations, and Design Insights,

    X. Lin, R. Wiren, S. Euler, A. Sadam, H.-L. M ¨a¨att¨anen, S. Muru- ganathan, S. Gao, Y .-P. E. Wang, J. Kauppi, Z. Zou, and V . Yaj- nanarayana, “Mobile Network-Connected Drones: Field Trials, Simu- lations, and Design Insights,” IEEE Vehicular Technology Magazine , vol. 14, ...

    2019

  7. [15]

    5G Aero: A Prototyping Platform for Evaluating Aerial 5G Communica- tions,

    M. Bordin, M. S. Holla, S. Velumani, S. D’Oro, and T. Melodia, “5G Aero: A Prototyping Platform for Evaluating Aerial 5G Communica- tions,” arXiv preprint arXiv:2506.08386 , 2025

    2025 arXiv

  8. [16]

    LTE I/Q Data Set for UA V Propagation Modeling, Communication, and Navigation Research,

    Maeng, Sung Joon and Ozdemir, Ozgur and Guvenc, Ismail and Sichitiu, Mihail L. and Mushi, Magreth and Dutta, Rudra, “LTE I/Q Data Set for UA V Propagation Modeling, Communication, and Navigation Research,” IEEE Communications Magazine , vol. 61, no. 9, pp. 90–96, 2023

    2023

  9. [17]

    AERIQ: SDR-Based LTE I/Q Measurement and Analysis Framework for Air-to-Ground Propagation Modeling,

    Maeng, S. J. and Ozdemir, O. and G ¨uvenc ¸,˙I. and Sichitiu, M. L. and Dutta, R. and Mushi, M., “AERIQ: SDR-Based LTE I/Q Measurement and Analysis Framework for Air-to-Ground Propagation Modeling,” in 2023 IEEE Aerospace Conference , 2023, pp. 1–11

    2023

  10. [18]

    Technical Feasibility and Design Challenges of Unmanned Aerial Vehicle based Drive Testing on Cellular Networks,

    Tak ´acs, Oliv ´er and Kov ´acs, Tam ´as and Drot ´ar, Istv ´an and Wers ´enyi, Gy¨orgy, “Technical Feasibility and Design Challenges of Unmanned Aerial Vehicle based Drive Testing on Cellular Networks,” in2022 IEEE 1st International Conference on Internet of Digital Reality (...

    2022

  11. [19]

    Nvidia, “Online,” Available https://www.nvidia.com/en-us/ autonomous-machines/embedded-systems/jetson-orin/

  12. [20]

    ROS, “Online,” Available https://www.ros.org/

  13. [21]

    pMLTE, “Online,” Available https://www.microhardcorp.com/pMLTE. php

  14. [22]

    Nping, “Online,” Available https://nmap.org/nping/

  15. [23]

    iPerf, “Online,” Available https://iperf.fr/

  16. [24]

    Networks, “Online,” Available https://wiki.teltonika-networks.com/ view/Mobile Signal Strength Recommendations

    T. Networks, “Online,” Available https://wiki.teltonika-networks.com/ view/Mobile Signal Strength Recommendations

  17. [25]

    OOKLA, “Online,” Available https://www.speedtest.net/

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.