Pith. sign in

REVIEW 3 major objections 4 minor 64 references

BatStation: Toward In-Situ Radar Sensing on 5G Base Stations with Zero-Shot Template Generation

T0 review · 3 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read This paper shows a 5G base station can detect, classify, and localize radar pulses from its own uplink resource grids using zero-shot templates generated from purely synthetic data, with no experimental fine-tuning.

desk verdict Solid systems paper with real OTA evidence, but the 'zero-shot' label is weaker than it claims because the detection threshold is fit to experimental no-radar grids and the cancellation assumption is never verified. read the letter →

arxiv 2509.06898 v1 pith:7PMNIQXH submitted 2025-09-08 cs.NI eess.SP

Zhihui Gao , Zhecun Liu , Tingjun Chen This is my paper
classification cs.NIeess.SP
keywords radarsensing5GbasestationCBRSspectrumsharingzero-shottemplategenerationsignalcancellationcorrelationuplinkresourcegridSDRtestbed
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

BatStation asks whether a 5G base station can act as an in-situ radar sensor in shared bands like CBRS, where naval radar and uplink 5G traffic occupy the same spectrum. The paper's central claim is that a radar pulse can be extracted from the base station's received resource grid by reconstructing and canceling the concurrent 5G uplink signal, then correlated against a small set of radar templates generated entirely from synthetic data. Because the templates are built from simulated radar waveforms rather than experimental recordings, the sensing model is hardware-portable and light enough to run in real time: 4,560 parameters, 40M MACs, and sub-millisecond latency. Experiments with real 5G traffic report mean detection probabilities of 97.02% on light-control uplink slots (PUCCH) and 79.23% on heavy-data uplink slots (PUSCH), classification accuracy up to 97.00%, and median localization errors of 2.68-6.20 MHz in frequency and 24.6-32.4 microseconds in time. If correct, this points to a path where spectrum sharing is driven by the base station itself rather than by dedicated radar sensors.

What carries the argument

The load-bearing object is the radar template set: a small collection of 2D patterns, one or two per radar type (for up- and down-chirps), encoding how each radar pulse appears on the reshaped 5G resource grid. Template correlation acts as a single 2D convolutional layer with no nonlinearities, so its outputs scale linearly with receiver gain, which is what makes the model portable across hardware and trainable on synthetic data alone. Supporting mechanisms are radar signal separation, which reconstructs and subtracts the 5G uplink signal using Hampel-filtered DMRS channel estimates and nearest-constellation rounding of data symbols, and resource grid reshaping, which uses a re-FFT to improv

What would settle it

Inject radar pulses into a live 5G uplink at INR 24-38 dB while logging the decoder's block error rate and the residual grid energy after BatStation's cancellation; if data symbols are frequently not demodulated correctly or the residual grid still correlates with scheduled 5G data, the separation step fails. A cleaner test: compare detection probability when the base station knows the true uplink data symbols (ideal cancellation) versus when it must round the constellation; a large gap would confirm the demodulation assumption is load-bearing.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the radar pattern visible on a 5G uplink resource grid is stable enough within a radar type to be captured by a single template, and that this template does not need to be learned from real radar recordings. BatStation constructs templates by simulating radar waveforms, projecting them through the same reshaping the received grid undergoes, averaging the resulting patterns per radar type, and fine-tuning on the synthetic corpus with a cross-entropy loss. At inference, sensing is a 2D correlation between the reshaped residual grid and the template set; the maximum over radar type, frequency, and time indexes gives detection, classification, and localizati

Load-bearing premise

Uplink data symbols that a radar pulse lands on can still be demodulated, or corrected by channel decoding, so the transmitted 5G constellation can be reconstructed and subtracted; otherwise residual 5G energy leaks into the correlation and breaks sensing.

Editorial extensions

If this is right

  • If the templates transfer zero-shot, a deployed base station can sense new radar types by extending the synthetic corpus, without collecting experimental radar data.
  • The sub-millisecond runtime (0.11 ms on GPU, 0.94 ms on CPU) is compatible with the 5 ms TDD periodicity, so sensing output can feed scheduling decisions in the same loop that allocates uplink and downlink slots.
  • The linearity of template correlation means the same templates can be applied on different base station hardware with no recalibration except the detection threshold.
  • Residual 5G leakage after cancellation behaves like extra interference: detection on data-heavy PUSCH slots is roughly 15 dB worse than on control-heavy PUCCH slots, so separation quality determines the sensing range.
  • Multiple radar pulses in one resource grid can be reported as multiple local maxima above threshold, extending the design from single-pulse detection to burst-level sensing without changing the model.

Reading between the lines

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

  • Editorial inference: if the decoder's error rate under radar interference is not negligible, the one-shot constellation rounding in the separation step will leak 5G energy; a natural variant would iterate between decoding and radar estimation.
  • Editorial inference: the same resource-grid reshaping plus template correlation recipe could generalize to other incumbent signals in shared bands, provided their time-frequency footprints can be simulated.
  • Editorial inference: the reported median frequency errors of up to 6-14 MHz on wideband chirps suggest that localization granularity is set by the pooled subcarrier spacing; finer localization would require smaller pooling factors or sub-grid interpolation at higher computational cost.
  • Editorial inference: because classification relies on learned negative regions in the templates, confusion among narrowband types under nonlinear channel effects such as clipping or Doppler is the first place portability could break; testing the templates on recorded radar data from diverse receivers would expose that.
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 / 4 minor

Summary. The paper presents BatStation, an in-situ radar sensing system for 5G base stations in shared spectrum (CBRS). It separates radar signals from concurrent 5G uplink transmissions by reconstructing and canceling the 5G signal from hard-decision demodulation, reshapes the residual time-frequency grid via re-FFT and max pooling, and uses radar templates initialized and fine-tuned on purely synthetic data. The system is evaluated on an SDR testbed with commodity smartphones and OTA radar signals, reporting detection probabilities of 97.02% (PUCCH) and 79.23% (PUSCH), classification accuracies of 97.00% and 95.30%, median frequency localization errors of 2.68–6.20 MHz, median time localization errors of 24.6–32.4 µs, and GPU/CPU latencies around 0.11/0.94 ms.

Significance. If the claims hold, the contribution is significant: it offers a standard-compatible, lightweight, in-situ radar sensing capability for 5G BSs without dedicated sensing hardware, with a linear template-correlation model having only 4,560 parameters. The OTA evaluation with real 5G traffic and the comparison against three baselines are strengths, and the paper is generally clear. However, the central 'true zero-shot' claim and the reported detection/localization numbers are not fully established because the detection threshold is calibrated on experimental no-radar data and because the signal-separation module rests on an unverified hard-decision demodulation assumption. These issues are load-bearing for the paper's main claims, though they appear addressable with additional measurements or a modified threshold-selection procedure.

major comments (3)
  1. [§7.2 and §5.3] The zero-shot claim is weakened by threshold selection in §7.2. Detection in Eq. (7) depends on Y_th, and Y_th is set to the 95th percentile of Y* from experimental no-radar resource grids, separately for PUCCH (28.8 dB) and PUSCH (35.1 dB). This is experimental calibration of the detector, so the end-to-end system is not 'true zero-shot' as stated in §5.3. All detection probabilities and downstream tasks depend on this threshold. Please either rephrase the claim to scope zero-shot to template generation only, or derive Y_th from synthetic/noise-only data (e.g., a CFAR-style rule) and re-evaluate.
  2. [§5.1, Eqs. (1)–(3)] The radar signal separation assumes interfered uplink data symbols can still be correctly demodulated or corrected by decoding, but the pipeline in Fig. 3 uses hard decisions directly and does not implement a decoder. Under INR of 24.3–38.4 dB, radar interference can push equalized symbols across decision boundaries; a wrong hard decision in Eq. (1) makes the subtraction in Eq. (3) inject residual 5G energy into Y_res, which can then correlate with radar templates. The reported aggregate INR reduction of 21.7–32.9 dB does not verify per-symbol decision correctness. Please report measured demodulation error rate or BLER under the OTA INR conditions, or independently validate cancellation quality against known 5G content.
  3. [§7.4 and §7.2] The localization evaluation appears to be conditioned on successful detection, but this is not stated explicitly. For PUSCH, P0N#1 is detected with only 8.95% probability (§7.2), yet a median normalized time error of 18.06% (90.3 µs) is reported for this type. If the localization statistics are computed only on the detected subset, they describe a small, non-representative set of pulses. Please state the conditioning explicitly and, ideally, report localization error over all test samples, treating undetected pulses as failures or providing a separate error floor.
minor comments (4)
  1. [§5.3] Typo: 'P3N#1' should be 'P0N#1'.
  2. [Abstract and §7.5] Latency numbers are inconsistent: the abstract says 0.11/0.94 ms on GPU/CPU, the introduction says 0.11/0.90 ms on CPU/GPU, and §7.5 reports 0.11/0.20 ms on GPUs and 0.94 ms on Intel Xeon. Please reconcile.
  3. [Fig. 10] The text refers to a 'log threshold' without specifying whether Y_th is applied to the log-domain or linear output. Clarify the scale and the false-alarm definition.
  4. [References] References [56] and [57] appear to be the same paper and should be merged.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: templates are trained on synthetic data and evaluated on OTA data; the disclosed threshold calibration and unverified demodulation assumption are correctness risks, not circular reductions.

full rationale

The paper's central claim is that radar templates are generated from purely synthetic data and then applied to over-the-air (OTA) resource grids. The template initialization and fine-tuning in §5.3 use only the synthetic dataset described in §6.2, and the OTA PUCCH/PUSCH datasets are explicitly held out as testing-only ('true zero-shot generation... require no fine-tuning on these experimental datasets'). The detection threshold in §7.2 is calibrated from experimental no-radar resource grids to achieve a 5% false-alarm rate, but this is a disclosed decision-rule calibration, not a fit of the radar templates, and it does not by construction force the reported detection probabilities. The radar waveform generator (NIST) is shared between the synthetic training set and the OTA radar emissions, which weakens the novelty of 'zero-shot' across radar waveform domains, but the OTA evaluation still requires real 5G interference, channel effects, and hardware behavior not present in training, so the result is not equivalent to its inputs by construction. The assumption in §5.1 that interfered uplink data symbols can be correctly demodulated or corrected by decoding is an unverified correctness risk—especially since the decoder is not part of the BatStation pipeline—but it is an assumption about system behavior, not a circular derivation. Self-citations (e.g., SWIRLS [16]) appear only in related work and are not load-bearing. No circular step can be exhibited from the paper's equations or citations.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The system contributes a signal-processing pipeline rather than a new physical theory; its claims rest on hand-tuned reshaping parameters, a calibrated detection threshold, and templates fitted to a synthetic dataset, not on an independently derived law.

free parameters (4)
  • Detection threshold Y_th = 28.8 dB (PUCCH), 35.1 dB (PUSCH)
    Chosen as the 95th percentile of Y* over experimental no-radar resource grids to enforce a 5% false alarm rate; the reported detection probabilities depend directly on this calibration (Section 7.2).
  • Fine-tuned template weights T^(r) = 4,560 parameters
    Learned via SGD cross-entropy minimization on the synthetic dataset; these are the convolution kernels used for detection, classification, and localization (Section 5.3).
  • Resource reshaping hyperparameters = re-FFT factor alpha, per-type pooling sizes P(r) in Table 3
    Chosen by hand so the pooled resolution matches each radar type's bandwidth and duration; they determine the correlation peak to frequency and time mapping (Sections 5.2-5.3).
  • Hampel filter sliding window length = 3 DMRS symbols
    Selected empirically as the smallest window giving a meaningful median; used for CSI estimation before 5G signal reconstruction (Section 5.1).
assumptions (5)
  • domain assumption Interfered uplink data symbols can be correctly demodulated or corrected by decoding, so the transmitted 5G constellation is recoverable for cancellation.
    Stated in Section 5.1; if false, the residual grid Y_res contains 5G leakage and the radar templates cannot be trusted.
  • domain assumption CSI estimated from non-interfered DMRS symbols (via the Hampel filter) is accurate and stable across nearby OFDM symbols.
    Used to reconstruct and cancel both data and DMRS symbols in Equations (1)-(3); no independent validation of CSI accuracy is provided.
  • ad hoc to paper Synthetic data generated with a frequency-flat channel, AWGN, and randomized center frequency and start time is representative enough for zero-shot transfer to OTA hardware.
    The zero-shot claim depends on this transfer; the experimental threshold calibration in Section 7.2 suggests the transfer is not perfect.
  • domain assumption Only the five NIST-defined CBRS radar types with their Table 2 parameter ranges need to be considered, and each 0.5 ms resource grid contains at most one radar pulse.
    From Table 2 and Section 6.2; limits the classification and correlation search space.
  • domain assumption Template correlation (2D convolution) captures the radar pattern, and the max over channels and templates yields correct detection, classification, and localization.
    This is the system's core model; it assumes the radar signature after separation is a scaled, shifted template pattern.

how reviews work

0 comments
Cite this review

Pith. "Pith review of BatStation: Toward In-Situ Radar Sensing on 5G Base Stations with Zero-Shot Template Generation." pith.science (2026). https://pith.science/paper/7PMNIQXH

@misc{pith2026250906898,
  author       = {Pith},
  title        = {Pith review of: BatStation: Toward In-Situ Radar Sensing on 5G Base Stations with Zero-Shot Template Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7PMNIQXH}},
  note         = {Machine review of arXiv:2509.06898}
}
read the original abstract

The coexistence between incumbent radar signals and commercial 5G signals necessitates a versatile and ubiquitous radar sensing for efficient and adaptive spectrum sharing. In this context, leveraging the densely deployed 5G base stations (BS) for radar sensing is particularly promising, offering both wide coverage and immediate feedback to 5G scheduling. However, the targeting radar signals are superimposed with concurrent 5G uplink transmissions received by the BS, and practical deployment also demands a lightweight, portable radar sensing model. This paper presents BatStation, a lightweight, in-situ radar sensing framework seamlessly integrated into 5G BSs. BatStation leverages uplink resource grids to extract radar signals through three key components: (i) radar signal separation to cancel concurrent 5G transmissions and reveal the radar signals, (ii) resource grid reshaping to align time-frequency resolution with radar pulse characteristics, and (iii) zero-shot template correlation based on a portable model trained purely on synthetic data that supports detection, classification, and localization of radar pulses without fine-tuning using experimental data. We implement BatStation on a software-defined radio (SDR) testbed and evaluate its performance with real 5G traffic in the CBRS band. Results show robust performance across diverse radar types, achieving detection probabilities of 97.02% (PUCCH) and 79.23% (PUSCH), classification accuracy up to 97.00%, and median localization errors of 2.68-6.20 MHz (frequency) and 24.6-32.4 microseconds (time). Notably, BatStation achieves this performance with a runtime latency of only 0.11/0.94 ms on GPU/CPU, meeting the real-time requirement of 5G networks.

Figures

Figures reproduced from arXiv: 2509.06898 by the authors.

Figure 1
Figure 1. In shared spectrum bands such as CBRS, the radar [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The 5G time/frequency structure: a resource grid is dimensioned by [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. BatStation’s workflow: first, the radar signals are separated from the raw resource grid Yraw; then, the separated resource grid Yres is reshaped by re-conducting FFT and different levels of max pooling along the frequency dimension; finally, the set of pooled resource grids Ypool is respectively correlated with a set of radar templates T, and the radar sensing results can be obtained by finding the maximum values 𝑌… view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: (Top) The per-subcarrier CSI estimated using nor￾mal and interfered DMRS symbols. (Bottom) The constella￾tions of normal and interfered data symbols. yielding the residual resource grid, denoted as Yres, with the radar signals Yradar and noises Ynoise only, as shown in…
Figure 5
Figure 5. Figure 5: Based on a synthetic dataset, the initialized (by [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 8
Figure 8. Figure 8: The 5G signal INR distribution from two users in the PUSCH dataset. database. We use the open-source 5G radio access network (RAN) stack, srsRAN version 24.04.0 (commit-51e44a642). 5G configuration. The established OTA 5G link between the BS and user(s) is centered at …
Figure 9
Figure 9. Figure 9: Example PUSCH resource grids with interfering 5G and radar signals before and after radar signal separation. [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: The output likelihood distributions with and with [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 15
Figure 15. Figure 15: The normalized frequency errors by BatStation’s center frequency localization under 30 dB SNR. accuracies based on a single radar pulse with 30 dB SNR reach up to 68.51/83.06% for DeepRadar and RadYOLO, re￾spectively, lower than BatStation of 98.37%. However, the clas…
Figure 13
Figure 13. Figure 13: The radar type classification accuracy comparison [PITH_FULL_IMAGE:figures/full_fig_p011_13.png]
Figure 14
Figure 14. Figure 14: The confusion matrices of BatStation’s classifi￾cation under 30 dB SNR on the PUCCH and PUSCH datasets. for the PUSCH shown in [PITH_FULL_IMAGE:figures/full_fig_p011_14.png]
Figure 16
Figure 16. Figure 16: The normalized time errors by BatStation’s start￾ing time localization under 30 dB SNR. PUCCH PUSCH INR INR [PITH_FULL_IMAGE:figures/full_fig_p012_16.png]
Figure 17
Figure 17. Figure 17: The normalized frequency/time median and 25/75- [PITH_FULL_IMAGE:figures/full_fig_p012_17.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

64 extracted references · 63 canonical work pages

  1. [1]

    Share Technote

    2023. Share Technote. www.sharetechnote.com

    work page 2023

  2. [2]

    3GPP Release 17

    2024. 3GPP Release 17. https://www.3gpp.org/specifications- technologies/releases/release-17

    work page 2024

  3. [3]

    iPerf - The ultimate speed test tool for TCP, UDP and SCTP

    2025. iPerf - The ultimate speed test tool for TCP, UDP and SCTP. https://iperf.fr/

    work page 2025

  4. [4]

    2025. Open5GS. https://github.com/open5gs

    work page 2025

  5. [5]

    Ghufran Baig, Ian Kash, Bozidar Radunovic, Thomas Karagiannis, and Lili Qiu. 2018. Interference management for unlicensed users in shared CBRS spectrum. InProc. ACM CoNEXT’18

    work page 2018

  6. [6]

    Luca Baldesi, Francesco Restuccia, and Tommaso Melodia. 2022. ChARM: NextG spectrum sharing through data-driven real-time O- RAN dynamic control. InProc. IEEE INFOCOM’22

    work page 2022

  7. [7]

    Richard Bell, Kyle Watson, Tianyi Hu, Isamu Poy, Fred Harris, and Dinesh Bharadia. 2023. Searchlight: An accurate, sensitive, and fast radio frequency energy detection system. InProc. IEEE MILCOM’23

    work page 2023

  8. [8]

    Das, Himanshu Gupta, and Petar M

    Arani Bhattacharya, Ayon Chakraborty, Samir R. Das, Himanshu Gupta, and Petar M. Djuric. 2020. Spectrum patrolling with crowd- sourced spectrum sensors.IEEE Trans. Cogn. Commun. Netw.6, 1 (Mar. 2020), 271–281

    work page 2020

Show all 64 references
  1. [9]

    Léon Bottou. 2010. Large-scale machine learning with stochastic gradient descent. InProc. Springer COMPSTAT’2010

    2010

  2. [10]

    Roberto Calvo-Palomino, Arani Bhattacharya, Gerome Bovet, and Domenico Giustiniano. 2020. Short: LSTM-based GNSS spoofing de- tection using low-cost spectrum sensors. InProc. IEEE WoWMoM’20

    2020

  3. [11]

    Raied Caromi, Michael Souryal, and Timothy A Hall. 2019. RF dataset of incumbent radar signals in the 3.5 GHz CBRS band.Journal of Research of the National Institute of Standards and Technology124 (2019), 1

    2019

  4. [12]

    Raied Caromi, Michael Souryal, and Wen-Bin Yang. 2018. Detection of incumbent radar in the 3.5 GHz CBRS band. InProc. IEEE GlobalSIP’18

    2018

  5. [13]

    Karyn Doke, Shamik Sarkar, Blessing Okoro, Danijela Cabric, and Mariya Zheleva. 2024. RadVIEW: Robust radar detection and charac- terization in high-noise regimes. InProc. IEEE DySPAN’24

    2024

  6. [14]

    Ettus Research. 2024. USRP N321 - Ettus Research Product Page. https://www.ettus.com/all-products/usrp-n321/ Accessed: 2025-03-23

    2024

  7. [15]

    Marco Fiore. 2023. Full network sensing: Architecting 6G beyond communications.IEEE Network37, 3 (May 2023), 232–239

    2023

  8. [16]

    Zhihui Gao, Yiran Chen, and Tingjun Chen. 2023. Swirls: Sniffing Wi-Fi using radios with low sampling rates. InProc. ACM MobiHoc’23

    2023

  9. [17]

    Nakul Garg, Aritrik Ghosh, and Nirupam Roy. 2024. Litefoot: Ultra- low-power localization using ambient cellular signals. InProc. ACM SenSys’24

    2024

  10. [18]

    Ushasi Ghosh, Azuka Chiejina, Nathan Stephenson, Vijay K Shah, Srini- vas Shakkottai, and Dinesh Bharadia. 2024. SPARC: Spatio-temporal adaptive resource control for multi-site spectrum management in nextG cellular networks. InProc. ACM CoNEXT’24

    2024

  11. [19]

    Jonny Kong, Qiang Xu, Zixiao Lu, Shivang Aggar- wal, Imran Khan, Yuanjie Li, Y

    Moinak Ghoshal, Z. Jonny Kong, Qiang Xu, Zixiao Lu, Shivang Aggar- wal, Imran Khan, Yuanjie Li, Y. Charlie Hu, and Dimitrios Koutsoniko- las. 2022. An in-depth study of uplink performance of 5G mmWave networks. InProc. ACM SIGCOMM Workshop on 5G and Beyond Net- work Measuremen...

    2022

  12. [20]

    Jonny Kong, Qiang Xu, Zixiao Lu, Shivang Aggar- wal, Imran Khan, Jiayi Meng, Yuanjie Li, Y

    Moinak Ghoshal, Z. Jonny Kong, Qiang Xu, Zixiao Lu, Shivang Aggar- wal, Imran Khan, Jiayi Meng, Yuanjie Li, Y. Charlie Hu, and Dimitrios Koutsonikolas. 2023. Can 5G mmWave enable edge-assisted real-time object detection for augmented reality?. InProc. IEEE MASCOTS’23

    2023

  13. [21]

    Junfeng Guan, Suraj Jog, Sohrab Madani, Ruochen Lu, Songbin Gong, Deepak Vasisht, and Haitham Hassanieh. 2022. Enabling IoT self- localization using ambient 5G mmWave signals. InProc. ACM SIG- COMM’22

    2022

  14. [22]

    Junfeng Guan, Arun Paidimarri, Alberto Valdes-Garcia, and Bod- hisatwa Sadhu. 2021. 3-D imaging using millimeter-wave 5G signal reflections.IEEE Trans. Microw. Theory Tech.69, 6 (2021), 2936–2948

    2021

  15. [23]

    Yeswanth Guddeti, Raghav Subbaraman, Moein Khazraee, Aaron Schul- man, and Dinesh Bharadia. 2019. SweepSense: Sensing 5 GHz in 5 milliseconds with low-cost radios. InProc. USENIX NSDI’19

    2019

  16. [24]

    Davidson

    Mohamed Hammouda, Rong Zheng, and Timothy N. Davidson. 2019. Learning-theoretic multi-channel spectrum sensing and access in full- duplex cognitive radio networks with unknown primary user activities. IEEE Trans. Netw. Sci. Eng.6, 4 (Oct. 2019), 885–897

    2019

  17. [25]

    Frank R Hampel. 1974. The influence curve and its role in robust estimation.J. Am. Stat. Assoc69, 346 (1974), 383–393

    1974

  18. [26]

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep residual learning for image recognition. InProc. IEEE CVPR’16

    2016

  19. [27]

    Aleksandar Ichkov, Alexander Wietfeld, Marina Petrova, and Ljiljana Simić. 2025. HBF MU-MIMO With interference-aware beam pair link allocation for beyond-5G mm-Wave networks.IEEE Trans. Mobile Comput.24, 5 (May 2025), 4248–4262

    2025

  20. [28]

    Chiasserini, Falko Dressler, Y

    Imran Khan, Moinak Ghoshal, Joana Angjo, Sigrid Dimce, Mushahid Hussain, Paniz Parastar, Yenchia Yu, Claudio Fiandrino, Charalam- pos Orfanidis, Shivang Aggarwal, Ana C Aguiar, Ozgu Alay, Carla F. Chiasserini, Falko Dressler, Y. Charlie Hu, Steven Y. Ko, Dimitrios Koutsonikola...

    2024

  21. [29]

    Woo-Hyun Ko, Ushasi Ghosh, Ujwal Dinesha, Raini Wu, Srinivas Shakkottai, and Dinesh Bharadia. 2024. EdgeRIC: Empowering real- time intelligent optimization and control in NextG cellular networks. InProc. USENIX NSDI’24

    2024

  22. [30]

    Cheng Li, Nathaniel Raymondi, Bin Xia, and Ashutosh Sabharwal

  23. [31]

    Zana Limani Fazliu, Francesco Malandrino, Carla Fabiana Chiasserini, and Alessandro Nordio. 2022. A belief propagation solution for beam coordination in mmWave vehicular networks.IEEE Trans. on Wireless Commun.21, 12 (Dec. 2022), 10809–10823

    2022

  24. [32]

    Yago Lizarribar, Roberto Calvo-Palomino, Alessio Scalingi, Giuseppe Santaromita, Gérôme Bovet, and Domenico Giustiniano. 2024. ORAN- Sense: Localizing non-cooperative transmitters with spectrum sensing and 5G O-RAN. InProc. IEEE INFOCOM’24

    2024

  25. [33]

    Zhicheng Luo, Qianyi Huang, Rui Wang, Hao Chen, Xiaofeng Tao, Guihai Chen, and Qian Zhang. 2022. WISE: Low-cost wide band spectrum sensing using UWB. InProc. ACM SenSys’22

    2022

  26. [34]

    Siddharth Mohan, Kshitiz Bansal, Manideep Dunna, and Dinesh Bhara- dia. 2023. Retroreflection with amplification for long range mmWave sensing. InProc. ACM mmNets’23

    2023

  27. [35]

    NIST. 2025. Simulated Radar Waveform Generator. https://github.com/ usnistgov/SimulatedRadarWaveformGenerator. GitHub Repository

    2025

  28. [36]

    Taekkyung Oh, Sangwook Bae, Junho Ahn, Yonghwa Lee, Tuan Dinh Hoang, Min Suk Kang, Nils Ole Tippenhauer, and Yongdae Kim. 2024. Enabling physical localization of uncooperative cellular devices. In Proc. ACM MobiCom’24

    2024

  29. [37]

    Paniz Parastar, Andra Lutu, Özgü Alay, Giuseppe Caso, and Diego Perino. 2023. Spotlight on 5G: Performance, device evolution and challenges from a mobile operator perspective. InIEEE INFOCOM’23

    2023

  30. [38]

    Jinbo Peng, Zhe Chen, Zheng Lin, Haoxuan Yuan, Zihan Fang, Lingzhong Bao, Zihang Song, Ying Li, Jing Ren, and Yue Gao. 2024. Sums: Sniffing unknown multiband signals under low sampling rates. IEEE Trans. Mobile Comput.(2024)

    2024

  31. [39]

    Jinbo Peng, Junwen Duan, Zheng Lin, Haoxuan Yuan, Yue Gao, and Zhe Chen. 2025. SigChord: Sniffing wide non-sparse multiband signals for terrestrial and non-terrestrial wireless networks.arXiv preprint arXiv:2504.06587(2025)

    2025 arXiv

  32. [40]

    Joseph Redmon, Santosh Divvala, Ross Girshick, and Ali Farhadi. 2016. You only look once: Unified, real-time object detection. InProc. IEEE CVPR’16

    2016

  33. [41]

    Guillem Reus-Muns, Pratheek S Upadhyaya, Utku Demir, Nathan Stephenson, Nasim Soltani, Vijay K Shah, and Kaushik R Chowdhury

  34. [42]

    2017.Procedures for laboratory testing of environ- mental sensing capability sensor devices

    Frank H Sanders. 2017.Procedures for laboratory testing of environ- mental sensing capability sensor devices. Technical Report. Institute for Telecommunication Sciences

    2017

  35. [43]

    Neagin Neasamoni Santhi, Davide Villa, Michele Polese, and Tom- maso Melodia. 2025. InterfO-RAN: Real-time in-band cellular uplink interference detection with GPU-accelerated dApps.arXiv preprint arXiv:2507.23177(2025)

    2025

  36. [44]

    Shamik Sarkar, Milind Buddhikot, Aniqua Baset, and Sneha Kumar Kasera. 2021. DeepRadar: A deep-learning-based environmental sens- ing capability sensor design for CBRS. InProc. ACM MobiCom’21

    2021

  37. [45]

    Shamik Sarkar, Dongning Guo, and Danijela Cabric. 2024. RadY- OLOLet: Radar detection and parameter estimation using YOLO and wavelet.IEEE Trans. Cogn. Commun. Netw.(2024)

    2024

  38. [46]

    Aron Schott, Enrico Tosi, Federico Chiariotti, and Ljiljana Simić. 2024. Impact of realistic GNSS location errors on REM-based beam manage- ment in urban millimeter-wave networks. InProc. IEEE DySPAN’24

    2024

  39. [47]

    Anish Shastri, Andres Garcia-Saavedra, and Paolo Casari. 2024. Indoor millimeter wave localization using multiple self-supervised tiny neural networks.IEEE Commun. Lett.28, 5 (May 2024), 1034–1038

    2024

  40. [48]

    Sladjan Sava Petrovic. 2024. 5G switch - Force 5G only. https://play. google.com/store/apps/details?id=com.sladjan.sava.petg&hl=en_US Accessed: 2025-03-23

    2024

  41. [49]

    Yiwen Song, Changhan Ge, Lili Qiu, and Yin Zhang. 2023. 2ACE: Spectral profile-driven multi-resolutional compressive sensing for mmWave channel estimation. InProc. ACM MobiHoc’23

    2023

  42. [50]

    srsRAN. 2025. srsRAN: Open source 5G software radio access network. https://github.com/srsran/srsran_project. GitHub Repository

    2025

  43. [51]

    Raghav Subbaraman, Kevin Mills, Aaron Schulman, and Dinesh Bhara- dia. 2023. Crescendo: Towards wideband, real-time, high-fidelity spec- trum sensing systems. InProc. ACM MobiCom’23

    2023

  44. [52]

    Harry Urkowitz. 1967. Energy detection of unknown deterministic signals.Proc. IEEE55, 4 (1967), 523–531

    1967

  45. [53]

    Daniel Uvaydov, Salvatore D’Oro, Francesco Restuccia, and Tommaso Melodia. 2021. DeepSense: Fast wideband spectrum sensing through real-time in-the-loop deep learning. InProc. IEEE INFOCOM’21

    2021

  46. [54]

    Yaxiong Xie and Kyle Jamieson. 2022. NG-scope: Fine-grained teleme- try for nextg cellular networks.Proc. ACM POMACS’22(2022)

    2022

  47. [55]

    Haotian Xu, Xiangtian Ma, Chengke Wang, Xiong Wang, Chenren Xu, Feng Gao, and Linghe Kong. 2022. A neural network approach for wireless spectrum anomaly detection in 5G-unlicensed network.CCF Trans. Pervasive Comput. Interact.4, 4 (Apr. 2022), 465–473

    2022

  48. [56]

    Yijing Zeng, Roberto Calvo-Palomino, Domenico Giustiniano, Gerome Bovet, and Suman Banerjee. 2023. Adaptive uplink data compres- sion in spectrum crowdsensing systems.IEEE/ACM Transactions on Networking31, 5 (2023), 2207–2221

    2023

  49. [57]

    Yijing Zeng, Roberto Calvo-Palomino, Domenico Giustiniano, Gerome Bovet, and Suman Banerjee. 2023. Adaptive uplink data compression in spectrum crowdsensing systems.IEEE/ACM Trans. Netw.31, 5 (Oct. 2023), 2207–2221

    2023

  50. [58]

    2023.WiFi/4G/5G Based Wireless Sensing: Theories, Applications and Future Directions

    Daqing Zhang, Kai Niu, Jie Xiong, Fusang Zhang, and Xuanzhi Wang. 2023.WiFi/4G/5G Based Wireless Sensing: Theories, Applications and Future Directions. Springer Nature Singapore, 387–417

    2023

  51. [59]

    Shuli Zhu, Lingkun Li, Xuyu Wang, Changcheng Liu, Yuqin Jiang, Zengwei Huo, Hua Chai, Jiqiang Liu, Dan Tao, and Ruipeng Gao

  52. [60]

    Frederik Jonathan Zumegen, Ish Kumar Jain, and Dinesh Bharadia

  53. [62]

    Experience: Large-scale cellular localization for pickup position recommendation at black-hole. InProc. ACM MobiCom’23

  54. [2022]

    Commun.70, 2 (Feb

    Outer bounds for a joint communicating radar (Comm-Radar): The uplink case.IEEE Trans. Commun.70, 2 (Feb. 2022), 1197–1213

    2022

  55. [2023]

    SenseORAN: O-RAN-based radar detection in the CBRS band. IEEE J. Sel. Areas Commun.42, 2 (2023), 326–338

    2023

  56. [2024]

    BeamArmor: Seamless anti-jamming in 5G cellular networks with MIMO null-steering. InProc. ACM HotMobile’24

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.