Vision-Based Autonomous MM-Wave Reflector Using ArUco-Driven Angle-of-Arrival Estimation

\\Ajinkya Umesh Mulik; Amod Ashtekar; Campbell Freeman; Josue Marroquin; Miles Dillon Lantz; Mohammed E Eltayeb; Nan Inzali

REVIEW 5 major objections 5 minor 21 references

Vision-Based Autonomous MM-Wave Reflector Using ArUco-Driven Angle-of-Arrival Estimation

T0 review · 5 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A camera on a motorized metal plate reads ArUco markers to keep 60 GHz signals flowing around obstacles.

desk verdict A real proof-of-concept for ArUco-steered passive mmWave reflectors, but the abstract inflates the results and the control law ignores the vertical axis, so the headline claims need to be re-scoped before this can be accepted. read the letter →

arxiv 2506.05195 v2 pith:TVQR4TQK submitted 2025-06-05 cs.CV

Josue Marroquin , Nan Inzali , Miles Dillon Lantz , Campbell Freeman , Amod Ashtekar , \Ajinkya Umesh Mulik , Mohammed E Eltayeb This is my paper

classification cs.CV

keywords millimeter-wavecommunicationspassivereflectorArUcomarkersangle-of-arrivalestimationvision-aidedbeamsteeringnon-line-of-sightcoverage60GHzautonomous

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 tries to establish that a low-cost autonomous reflector can restore millimeter-wave links when the direct path is blocked. The core idea is to put ArUco markers on the transmitter and receiver, let a camera mounted on a motorized metal plate measure where each node is, and steer the plate so its surface normal bisects the incoming and outgoing rays. Indoor 60 GHz experiments report received-power gains up to 17 dB (about 10 dB on average) over the no-reflector baseline, and link availability rises from essentially zero to more than half at a -75 dB threshold. If correct, this gives a cheap, infrastructure-free middle path between static passive reflectors and electronically complex reconfigurable intelligent surfaces.

What carries the argument

The load-bearing mechanism is specular bisection of the incident and reflected rays. The camera maps each marker centroid's pixel coordinate $x$ to a bearing through $\theta_{\mathrm{raw}} = (x/W)\Theta_{\mathrm{FOV}}$ with an empirical barrel-distortion correction, then commands the reflector to $\theta_{\mathrm{ref}} = (\theta_T + \theta_R)/2$. This makes the incident and reflected angles equal, driving the bistatic angle $\beta$ toward zero, which the flat-plate radar cross section $4\pi A^2/\lambda^2 \cos(\beta/2)$ says maximizes reradiated power. The stepper's $0.1125^\circ$ resolution converts the angular command into a step count.

What would settle it

Raise the receiver 30 cm above the plane containing the camera and transmitter while keeping both markers visible, then measure received power at several horizontal offsets: if the vision-guided configuration no longer beats the no-reflector baseline by roughly 10 dB, the horizontal-only angle-of-arrival model is the cause.

Watch

Extended reading notes

Core claim

The paper's central claim is that monocular vision plus fiducial markers can replace RF feedback for reflector alignment. Received power over a flat conducting plate is maximized when the reflector normal bisects the two node bearings, and those bearings can be estimated from the horizontal pixel coordinate of each detected marker. A 300 mm square aluminum plate on a 0.9° stepper with 1/16 microstepping, driven by a Raspberry Pi and a global-shutter camera, recovered most of the gain of a manually aimed reflector: it stayed within 1–3 dB of manual alignment, beat the no-reflector baseline by an average of about 10 dB, and raised link availability at a -75 dB threshold from essentially 0% to about 53%. The abstract states a larger 23 dB average gain and 0.89 probability above -65 dB, while the body reports the more conservative figures; either way, the claimed contribution is the same: passive reflection can be steered by cheap vision alone.

Load-bearing premise

The alignment chain uses only the horizontal pixel position of each marker, so the transmitter, receiver, and camera are assumed to lie at the same height; if a node sits noticeably above or below that plane, the reflected beam will miss the receiver.

Editorial extensions

If this is right

Reflector alignment no longer needs RF probing or channel estimation, so NLoS coverage can be restored with a camera, a stepper, and a metal plate.
Because ArUco markers carry unique IDs, the reflector can serve only authenticated nodes, which reduces unintended signal exposure compared to broadcasting toward a blocked area.
If no marker is detected, the system falls back to a static reflector, so the deployment degrades gracefully under occlusion.
The 1–3 dB gap to a manually aimed reflector implies that vision steering captures most of the available passive gain, leaving little performance on the table for autonomy.
The prototype's gains point to practical short-range uses in 60 GHz Wi-Fi, AR/VR headsets, and IoT where heavy RIS infrastructure is not available.

Reading between the lines

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

A natural next step is vertical-axis steering: the pipeline reads only the $x$-pixel coordinate, so it assumes all nodes sit at camera height, and the on-board two-axis gimbal already provides the hardware to fix this.
The marker-ID selectivity also introduces a security question the paper does not pursue: a printed marker can be spoofed, so genuinely selective service would need authenticated or signed fiducials.
Because the receiver antenna was omnidirectional, the measured gains are a conservative floor; coordinating reflector steering with a directional receiver beam would likely close more of the gap to the LoS baseline.
The distortion correction is an empirical fit to one lens; replacing it with a standard calibrated camera model would likely shrink the off-axis angle errors (up to about 5.7° in the table) below typical mmWave beamwidths with less tuning.

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.

Desk Editor's Note

read the letter

Two things you should know: this is a genuinely useful systems paper, not a conceptual breakthrough, and the "23 dB" in the abstract is contradicted by the full text's ~10 dB, which should not survive review. The genuinely new bit is combining ArUco marker angle-of-arrival estimation with a motorized passive reflector for 60 GHz NLoS coverage, plus selective alignment to authenticated markers. Prior vision-aided RIS work and static reflectors exist, but not this integrated package. The authors built a real Raspberry Pi prototype, ran indoor measurements against three baselines, and were honest enough to report that vision steering lands within 1–3 dB of manual alignment and lags LoS by ~9–11 dB. The LoS path-loss regression (n = 1.57, R² = 0.981) and the AoA error table give the system credibility. That is real engineering evidence.

The soft spots are real but not fatal. First, the abstract numbers are unsupported; the body reports up to 17 dB and average ≈10 dB, with no confidence intervals, no trial counts, and no stated node heights. Second, the stress-test concern is correct: Algorithm 1 uses only the marker x-pixel coordinate and commands a single azimuth angle. The two-axis gimbal is mentioned but the elevation axis is never used. Specular reflection is a 3D problem; a 10 cm height mismatch at 2 m is about 2.9°, larger than the roughly 0.95° reflected beamwidth for a 300 mm plate at 60 GHz. Unless nodes happened to be coplanar with the camera, the measured gains will not reproduce. The paper must either add vertical steering or explicitly scope the claims to height-coplanar nodes. Third, "dynamic" overstates the experiment: the receiver was moved laterally in 0.5 ft increments, which is static re-pointing, not tracking a moving node.

On balance, the central idea holds as a proof of concept for the coplanar case, and the limitations section lists several practical constraints, which is more than many systems papers do. The citation pattern looks appropriate; the self-citation is to the authors' own LiDAR-aided reflector work, which is relevant.

Who this is for: people building low-cost 60 GHz coverage solutions or vision-aided beam management. It deserves a serious referee, but the review must demand a reconciled abstract, trial counts and heights, and a clear statement of the planar limitation. If the authors fix those, this is a solid venue-level contribution.

Referee Report

5 major / 5 minor

Summary. The manuscript describes and experimentally evaluates a low-cost, vision-guided passive reflector for 60 GHz NLoS links. A Raspberry Pi detects ArUco markers on the transmitter and receiver with a monocular camera, estimates their horizontal angles of arrival, and rotates a motorized metallic plate to bisect the estimated directions. The full-text experiments compare the vision-guided reflector against a no-reflector NLoS baseline, a manually aimed reflector, and a LoS baseline, reporting up to 17 dB improvement (average ≈10 dB) and an outage improvement from essentially 0% to ≈53% availability at a −75 dB threshold. The abstract, however, claims a 23 dB average gain and 0.89 availability above −65 dB, which do not appear in the body.

Significance. The core idea is appealing: replacing RF feedback with camera-based fiducial tracking makes the reflector infrastructure-free and adds user identification via unique markers. Strengths include a complete hardware description, direct received-power measurements, and a LoS path-loss regression (n=1.57, R²=0.981) that anchors the measurement methodology. If the claims are narrowed to horizontal steering of coplanar, quasi-static nodes, the work is a useful proof-of-concept. The current text, however, extends the claims to dynamic 3-D scenarios and reports unsupported headline numbers, so the significance as stated is not yet established.

major comments (5)

[Abstract and Sections V–VI] The Abstract reports a 23 dB average gain and a 0.89 probability of maintaining reception above −65 dB, but Sections V and VI report a maximum improvement of 17 dB, an average gain of ≈10 dB, and roughly 53% link availability at a −75 dB threshold. These are different claims; the abstract numbers are not supported by any figure, table, or trial description in the body. The authors should either supply the measurements behind the abstract numbers or revise the abstract to match the full text.
[Algorithm 1 and Section III-C] The control law uses only the horizontal pixel coordinate x_i (line 4 and Eq. (3)) and commands a single scalar reflector orientation θref = (θT+θR)/2 (Eq. (7)). The vertical pixel coordinate and the elevation axis of the two-axis gimbal are never used, so the system does not control the reflected ray in elevation. Specular reflection in 3-D requires the plate normal to bisect the incident and reflected rays in elevation as well as azimuth; with the 300 mm plate at 60 GHz (mainlobe half-width ≈0.95°), a height mismatch of only 10 cm at 2 m (≈2.9°) is enough to move the receiver out of the reflected beam. Since the Section IV experiment moved the receiver only laterally and does not report node or camera heights, the central claim of autonomous 3-D reflector steering is not supported by the reported data.
[Section V, Figs. 6–7] The central empirical claims—average ≈10 dB gain and outage improvement—are presented as mean values and CCDFs without the number of trials, confidence intervals, or error bars. The path-loss baseline reports an R² but the reflector comparisons do not report any variability. Without trial counts and spread, the reader cannot determine whether the vision-guided reflector's ≈53% availability is statistically distinguishable from the manually aimed case (≈63%) or whether the ≈10 dB average gain is robust.
[Sections IV–V and Abstract] The abstract and introduction describe operation in 'dynamic' NLoS scenarios, but the experimental protocol places the receiver at discrete lateral positions and measures power at each point; no continuous motion, tracking latency, occlusion, or live mobility scenario is reported. The phrase 'real time' is also not quantified (no frame rate or control-loop latency is given). The dynamic-operation claim should be removed or supported by a dedicated dynamic experiment.
[Section III-A, Eq. (4), Table I] The distortion correction in Eq. (4) is introduced as an 'empirically tuned transformation' with constants 43.5 and 4000, but no fitting data, calibration procedure, or validation set is provided. Because this correction directly determines the AoA used in Eq. (7), and Table I reports errors of 3–6° at off-axis angles, the impact of these AoA errors on the reflected beam alignment should be quantified. At the stated plate beamwidth of ≈0.95°, such errors could place the receiver in the first null; the paper needs to explain why the link-level measurements remain in good agreement despite these errors.

minor comments (5)

[Section IV] The text says the motor is 0.9° with 1/16 microstepping and reports Δθ=0.1125°, but 0.9/16 = 0.05625°; either the step angle or the resolution should be corrected.
[Equation (4)] The units are unclear because θraw and θc are angles while the denominator 4000 appears to be a pixel count; specify all units and how the constants were obtained.
[Sections I and II] The user-identification/security benefit of ArUco markers is claimed in Section I but never evaluated experimentally; a brief statement that this is a qualitative feature would be appropriate.
[Table I] It would help to report the elevation/vertical position of the marker relative to the camera, since the algorithm ignores the y-coordinate and the coplanarity assumption is central to the result.
[Section VI] The conclusion mentions the omnidirectional RX as a conservative lower bound; if a directional RX was available, showing even one comparison would strengthen the claim, but as written this is a limitation to acknowledge.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the central results are direct experimental measurements, with calibration fits and a background self-citation that are not load-bearing.

full rationale

The paper's central claims are measured received-power comparisons across four configurations (LoS, no reflector, manually aligned reflector, and vision-guided reflector), not predictions from a fitted model. The LoS path-loss regression (a = -56.10 dB, b = -15.71 dB/decade, n = 1.57) is used only to validate the measurement methodology against independent indoor mmWave path-loss exponents, not to produce the reflector gains. Equation (4) is an empirically tuned distortion correction, but its coefficients are fixed calibration inputs; the AoA validation in Table I compares the corrected estimates to independent protractor measurements, so the error report is not circular. Equation (7) is simply the Snell's-law bisector rule, a physical constraint rather than an imported uniqueness claim. The only self-citation, Ref. [14] on LiDAR-aided passive mirrors, appears in background context and is not load-bearing for the results. The paper's limitation regarding vertical/elevation control is a scope and correctness issue, not a circularity issue. Overall, the derivation chain is self-contained with respect to its experimental claims, and no step reduces by construction to its own inputs.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The experimental claim is direct, so the ledger is small. The only fitted numbers are the two constants in Eq. (4). The analysis relies on standard radar and optics formulas plus the coplanarity assumption hidden in the horizontal-only angle computation.

free parameters (2)

distortion scale factor = 43.5
Empirical constant in Eq. (4) shaping the pixel-to-angle correction curve; chosen by hand with no calibration procedure stated.
distortion denominator = 4000
Empirical constant in Eq. (4) controlling angular scale; fitted to the camera lens, not derived.

assumptions (4)

domain assumption The bistatic radar equation (Eq. 1) applies in far-field conditions
Used to justify received power dependence on distances and RCS, assuming no near-field or multipath effects.
domain assumption The perfectly conducting flat plate RCS formula (Eq. 2) approximates the reflector
Assumes specular reflection and ignores edge diffraction and material losses.
domain assumption Snell's law: the reflector normal bisects the incident and reflected angles
Sets θ_ref = (θ_T + θ_R)/2, assuming a 2D geometry where all nodes lie in the camera's horizontal plane.
domain assumption ArUco marker detection reliably provides pixel centroids
The whole pipeline depends on OpenCV's ArUco detector, with no formal guarantee under occlusion or varying lighting.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Vision-Based Autonomous MM-Wave Reflector Using ArUco-Driven Angle-of-Arrival Estimation." pith.science (2026). https://pith.science/paper/TVQR4TQK

@misc{pith2026250605195,
  author       = {Pith},
  title        = {Pith review of: Vision-Based Autonomous MM-Wave Reflector Using ArUco-Driven Angle-of-Arrival Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TVQR4TQK}},
  note         = {Machine review of arXiv:2506.05195}
}

read the original abstract

Reliable millimeter-wave (mmWave) communication in non-line-of-sight (NLoS) conditions remains a major challenge for both military and civilian operations, especially in urban or infrastructure-limited environments. This paper presents a vision-aided autonomous reflector system designed to enhance mmWave link performance by dynamically steering signal reflections using a motorized metallic plate. The proposed system leverages a monocular camera to detect ArUco markers on allied transmitter and receiver nodes, estimate their angles of arrival, and align the reflector in real time for optimal signal redirection. This approach enables selective beam coverage by serving only authenticated targets with visible markers and reduces the risk of unintended signal exposure. The designed prototype, built on a Raspberry Pi 4 and low-power hardware, operates autonomously without reliance on external infrastructure or GPS. Experimental results at 60\,GHz demonstrate a 23\,dB average gain in received signal strength and an 0.89 probability of maintaining signal reception above a target threshold of -65 dB in an indoor environment, far exceeding the static and no-reflector baselines. These results demonstrate the system's potential for resilient and adaptive mmWave connectivity in complex and dynamic environments.

Figures

Figures reproduced from arXiv: 2506.05195 by the authors.

**Figure 2.** Camera view from the reflector system. ArUco markers [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** Prototype architecture. A Raspberry Pi captures cam [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Measurement and environment setup illustrating the [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

**Figure 6.** Figure 6: Complementary CDF of the received relative power [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]

**Figure 7.** Figure 7: Performance comparison: (a) Average received relative [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

21 extracted references · 20 canonical work pages

[1]

Terahertz communications and sensing for 6G and beyond: A compre- hensive review,

W. Jiang, B. Shen, L. Zhang, Y . Zeng, M. Shafi, and T. S. Rappaport, “Terahertz communications and sensing for 6G and beyond: A compre- hensive review,” IEEE Communications Surveys & Tutorials , vol. 26, no. 4, pp. 2326–2381, Fourthquarter 2024

work page 2024
[2]

Exploiting high millimeter wave bands for military communications, applications, and design,

J. F. Harvey, M. B. Steer, and T. S. Rappaport, “Exploiting high millimeter wave bands for military communications, applications, and design,” IEEE Access, vol. 7, pp. 52350–52359, 2019

work page 2019
[3]

Joint Beam Management and Relay Selection Using Deep Reinforcement Learning for MmWave UA V Relay Networks,

D. Kim, M. R. Castellanos, and R. W. Heath, “Joint Beam Management and Relay Selection Using Deep Reinforcement Learning for MmWave UA V Relay Networks,” in MILCOM 2022 - 2022 IEEE Military Communications Conference (MILCOM) , Rockville, MD, USA, 2022, pp. 895–900

work page 2022
[4]

Tracking and Identification of Targets via mmWave MIMO Radar,

A. Vaccari, M. Z. Win, and A. Conti, “Tracking and Identification of Targets via mmWave MIMO Radar,” in MILCOM 2024 - 2024 IEEE Military Communications Conference (MILCOM) , Washington, DC, USA, 2024, pp. 336–341

work page 2024
[5]

Liu et al., “Reconfigurable intelligent surfaces: Principles and opportunities,“IEEE Commun

Y . Liu et al., “Reconfigurable intelligent surfaces: Principles and opportunities,“IEEE Commun. Surveys Tuts. , vol. 23, no. 3, pp. 1546– 1577, 2021

work page 2021
[6]

RIS-aided Wireless Communications: Prototyping, Adap- tive Beamforming, and Indoor/Outdoor Field Trials,

X. Pei et al., “RIS-aided Wireless Communications: Prototyping, Adap- tive Beamforming, and Indoor/Outdoor Field Trials,” IEEE Trans. Commun., vol. 69, no. 12, pp. 8627–8640, Dec. 2021

work page 2021
[7]

Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,

M. A. ElMossallamy et al., “Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,” IEEE Trans. Cogn. Commun. Netw., vol. 6, no. 3, pp. 990–1002, 2020

work page 2020
[8]

Computer Vision-Aided Reconfigurable Intelligent Surface-Based Beam Tracking: Prototyping and Experimental Results,

M. Ouyang, Y . Wang, F. Gao, S. Zhang, P. Li, and J. Ren, “Computer Vision-Aided Reconfigurable Intelligent Surface-Based Beam Tracking: Prototyping and Experimental Results,” IEEE Transactions on Wireless Communications, vol. 22, no. 12, pp. 8681–8693, Dec. 2023

work page 2023

Show all 21 references

[9]

Camera Aided Reconfigurable Intelligent Surfaces: Computer Vision Based Fast Beam Selection,

S. Jiang, A. Hindy, and A. Alkhateeb, “Camera Aided Reconfigurable Intelligent Surfaces: Computer Vision Based Fast Beam Selection,” in Proceedings of the IEEE International Conference on Communications (ICC), Rome, Italy, May 2023, pp. 2921–2926

2023
[10]

Vision-Aided 6G Wireless Communications: Block- age Prediction and Proactive Handoff,

G. Charan et al., “Vision-Aided 6G Wireless Communications: Block- age Prediction and Proactive Handoff,” IEEE Trans. Veh. Technol., vol. 70, no. 10, pp. 10193-10208, 2021

2021
[11]

Machine learning on camera images for fast mmwave beamforming,

B. Salehi et al., “Machine learning on camera images for fast mmwave beamforming,” in Proc. IEEE MASS , 2020, pp. 338–346

2020
[12]

Coverage enhancement for NLOS mmWave links using passive reflectors,

W. Khawaja, O. Ozdemir, Y . Yapici, F. Erden, M. Ezuma, and I. Guvenc, “Coverage enhancement for NLOS mmWave links using passive reflectors,” IEEE Open Journal of the Communications Society , vol. 1, pp. 263–281, 2020

2020
[13]

Effect of passive reflectors for enhancing coverage of 28 GHz mmWave systems in an outdoor setting,

W. Khawaja, O. Ozdemir, F. Erden, I. Guvenc, M. Ezuma, and Y . Kakishima, “Effect of passive reflectors for enhancing coverage of 28 GHz mmWave systems in an outdoor setting,” in Proc. IEEE Radio and Wireless Symposium (RWS), Orlando, FL, USA, Jan. 2019, pp. 1–3

2019
[14]

LiDAR-Aided Millimeter-Wave Range Extension Using a Passive Mirror Reflector,

O. Ibrahim, R. S. S. Bandari, and M. E. Eltayeb, “LiDAR-Aided Millimeter-Wave Range Extension Using a Passive Mirror Reflector,” in Proc. 2025 IEEE 22nd Consumer Communications & Networking Conference (CCNC), Las Vegas, NV , USA, 2025, pp. 1–4

2025
[15]

DeepArUco++: Improved detection of square fiducial markers in challenging lighting conditions,

R. Berral-Soler, R. Mu ˜noz-Salinas, R. Medina-Carnicer, and M. J. Mar ´ın-Jim´enez, “DeepArUco++: Improved detection of square fiducial markers in challenging lighting conditions,” Image and Vision Computing, vol. 145, p. 105313, Oct. 2024

2024
[16]

Automatic generation and detection of highly reliable fiducial markers under occlusion,

S. Garrido-Jurado, R. Mu ˜noz-Salinas, F. J. Madrid-Cuevas, and M. J. Mar ´ın-Jim´enez, “Automatic generation and detection of highly reliable fiducial markers under occlusion,” Pattern Recognition, vol. 47, no. 6, pp. 2280–2292, June 2014

2014
[17]

T. S. Rappaport, Wireless Communications: Principles and Practice , 2nd ed. Upper Saddle River, NJ, USA: Prentice Hall, 2002

2002
[18]

On the derivation of bistatic RCS from monostatic mea- surements,

R.E. Kell, “On the derivation of bistatic RCS from monostatic mea- surements,” Proc. IEEE, V ol.53, pp983-988, 1965

1965
[19]

ArUco Marker Detection,

OpenCV , “ArUco Marker Detection,” OpenCV Documen- tation, Accessed: May 30, 2025. [Online]. Available: https://docs.opencv.org/4.x/d5/dae/utorial aruco detection.html

2025
[20]

Indoor office wideband millimeter wave propagation measurements and chan- nel models at 28 and 73 GHz,

G. R. MacCartney, S. Sun, T. S. Rappaport, and S. Rangan, “Indoor office wideband millimeter wave propagation measurements and chan- nel models at 28 and 73 GHz,” IEEE Access , vol. 3, pp. 2388–2424, 2015, doi: 10.1109/ACCESS.2015.2486778

2015
[21]

Propagation path loss models for 5G urban micro and indoor office scenarios,

S. Sun, T. S. Rappaport, G. R. MacCartney, M. K. Samimi, and S. Rangan, “Propagation path loss models for 5G urban micro and indoor office scenarios,” in Proc. IEEE 81st Veh. Technol. Conf. (VTC Spring), 2015, pp. 1–6, doi: 10.1109/VTCSpring.2015.7146039

2015

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Terahertz communications and sensing for 6G and beyond: A compre- hensive review,

W. Jiang, B. Shen, L. Zhang, Y . Zeng, M. Shafi, and T. S. Rappaport, “Terahertz communications and sensing for 6G and beyond: A compre- hensive review,” IEEE Communications Surveys & Tutorials , vol. 26, no. 4, pp. 2326–2381, Fourthquarter 2024

work page 2024

[2] [2]

Exploiting high millimeter wave bands for military communications, applications, and design,

J. F. Harvey, M. B. Steer, and T. S. Rappaport, “Exploiting high millimeter wave bands for military communications, applications, and design,” IEEE Access, vol. 7, pp. 52350–52359, 2019

work page 2019

[3] [3]

Joint Beam Management and Relay Selection Using Deep Reinforcement Learning for MmWave UA V Relay Networks,

D. Kim, M. R. Castellanos, and R. W. Heath, “Joint Beam Management and Relay Selection Using Deep Reinforcement Learning for MmWave UA V Relay Networks,” in MILCOM 2022 - 2022 IEEE Military Communications Conference (MILCOM) , Rockville, MD, USA, 2022, pp. 895–900

work page 2022

[4] [4]

Tracking and Identification of Targets via mmWave MIMO Radar,

A. Vaccari, M. Z. Win, and A. Conti, “Tracking and Identification of Targets via mmWave MIMO Radar,” in MILCOM 2024 - 2024 IEEE Military Communications Conference (MILCOM) , Washington, DC, USA, 2024, pp. 336–341

work page 2024

[5] [5]

Liu et al., “Reconfigurable intelligent surfaces: Principles and opportunities,“IEEE Commun

Y . Liu et al., “Reconfigurable intelligent surfaces: Principles and opportunities,“IEEE Commun. Surveys Tuts. , vol. 23, no. 3, pp. 1546– 1577, 2021

work page 2021

[6] [6]

RIS-aided Wireless Communications: Prototyping, Adap- tive Beamforming, and Indoor/Outdoor Field Trials,

X. Pei et al., “RIS-aided Wireless Communications: Prototyping, Adap- tive Beamforming, and Indoor/Outdoor Field Trials,” IEEE Trans. Commun., vol. 69, no. 12, pp. 8627–8640, Dec. 2021

work page 2021

[7] [7]

Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,

M. A. ElMossallamy et al., “Reconfigurable intelligent surfaces for wireless communications: Principles, challenges, and opportunities,” IEEE Trans. Cogn. Commun. Netw., vol. 6, no. 3, pp. 990–1002, 2020

work page 2020

[8] [8]

Computer Vision-Aided Reconfigurable Intelligent Surface-Based Beam Tracking: Prototyping and Experimental Results,

M. Ouyang, Y . Wang, F. Gao, S. Zhang, P. Li, and J. Ren, “Computer Vision-Aided Reconfigurable Intelligent Surface-Based Beam Tracking: Prototyping and Experimental Results,” IEEE Transactions on Wireless Communications, vol. 22, no. 12, pp. 8681–8693, Dec. 2023

work page 2023