pith. sign in

arxiv: 2509.21290 · v3 · pith:TPYDHPHUnew · submitted 2025-09-25 · 📡 eess.SP

Vision-Intelligence-Enabled Beam Tracking for Cross-Interface Optical Wireless Communication between Underwater and Low-Altitude Platforms

Jiayue Liu , Tianqi Mao , Leyu Cao , Weijie Liu , Dezhi Zheng , Julian Cheng , Zhaocheng Wang This is my paper

Pith reviewed 2026-05-21 22:10 UTC · model grok-4.3

classification 📡 eess.SP
keywords optical wireless communicationbeam trackingwater-air interfacevision-based algorithmunderwater communicationsea surface refractionneural network
0
0 comments X

The pith

A vision-based neural algorithm tracks optical beams across the moving sea surface to sustain links between underwater devices and low-altitude platforms.

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

The paper models how light beams refract and misalign when crossing a time-varying water-air interface, creating a challenge for high-speed optical wireless communication. It develops a tracking method that feeds visual observations into a convolutional neural network followed by a bidirectional LSTM with attention to extract features and predict beam positions in real time. This targets the need for broadband data transfer in ocean surveillance and exploration, where acoustic links lack sufficient capacity. Simulations show the approach holds higher received signal strength and reduces noise effects compared with conventional tracking techniques.

Core claim

The paper establishes a mathematical channel model for water-air optical transmission across a time-varying sea surface. Based on the model, a vision-based beam tracking algorithm combining convolutional neural network and bi-directional long short-term memory with an attention mechanism is developed to extract key spatio-temporal features. Simulations verify that the proposed algorithm outperforms classical methods in maintaining received signal strength and suppressing vision noise, demonstrating its robustness for water-air OWC systems.

What carries the argument

The vision-based beam tracking algorithm that combines a convolutional neural network with bi-directional long short-term memory and an attention mechanism to extract spatio-temporal features from visual inputs for real-time beam alignment.

If this is right

  • The algorithm maintains higher received signal strength than classical methods under simulated dynamic conditions.
  • It suppresses vision noise more effectively than prior approaches.
  • It supports real-time transceiver alignment that adapts to complex oceanic dynamics.
  • The method demonstrates robustness for water-air optical wireless communication systems.

Where Pith is reading between the lines

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

  • The approach could reduce reliance on narrowband acoustic backhaul for large observational datasets from underwater sensors.
  • Integration with low-altitude platforms might enable continuous high-bandwidth links for extended mineral exploration or surveillance missions.
  • Adding wave-height or wind data as extra inputs could further improve prediction accuracy in future versions.

Load-bearing premise

The mathematical channel model for time-varying sea surface refraction and the simulated vision noise accurately represent real oceanic dynamics sufficiently to support the performance claims.

What would settle it

Real-sea experiments that compare measured beam misalignment and signal strength under actual wave conditions against the simulation predictions for the same sea states.

Figures

Figures reproduced from arXiv: 2509.21290 by Dezhi Zheng, Jiayue Liu, Julian Cheng, Leyu Cao, Tianqi Mao, Weijie Liu, Zhaocheng Wang.

Figure 1
Figure 1. Figure 1: Water-air OWC channel Model. the optical signal is emitted by laser diode (LD) at position T : (xt, yt, zt), crossing the water-air interface at position S : (xs, ys, zs), and received by avalanche photodiode (APD) at R : (xr, yr, zr). Note that the beam emitted from the LD exhibit strong directionality, making transceiver alignment extremely crucial for stable communication link [22]. To facilitate releva… view at source ↗
Figure 2
Figure 2. Figure 2: Oceanic surface illustrations based on the wave spectrum theory, including: (a) 3-dimensional ocean wave spectrum. (b) Wave surface simulated by [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: System model of the water-to-air OWC system within oceanic environment. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Model structure of the proposed AI Network. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Training History, including training/validation loss and learning rate [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Testing result of experimental training between the proposed algo [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Received signal strength under different optical beam alignment [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗
Figure 10
Figure 10. Figure 10: Simulated temporal communication BER. highlights the enhanced capabilities of the proposed algorithm compared to its counterpart. The superior performance of the proposed algorithm might be attributed to the visual scheme integrated with the algorith￾mic structure, which enables more precise sensing of the beam direction and aligns the transceiver orientation closer to the minimum OPL path. This configura… view at source ↗
read the original abstract

The rapid expansion of oceanic applications such as underwater surveillance and mineral exploration is driving the need for real-time wireless backhaul of massive observational data. Such demands are challenging to meet using the narrowband acoustic approach. Alternatively, with the participation of low-altitude platforms (LAPs), water-air optical wireless communication (OWC) has emerged as a promising solution owing to its high potential for broadband transmission. However, implementing water-air OWC remains challenging, particularly when signals penetrate the fluctuating interface, where dynamic refraction induces severe beam misalignment with airborne stations. This necessitates real-time transceiver alignment capable of adapting to complex oceanic dynamics, which remains largely unaddressed. Against this background, this paper establishes a mathematical channel model for water-air optical transmission across a time-varying sea surface. Based on the model, a vision-based beam tracking algorithm combining convolutional neural network and bi-directional long short-term memory with an attention mechanism is developed to extract key spatio-temporal features. Simulations verify that the proposed algorithm outperforms classical methods in maintaining received signal strength and suppressing vision noise, demonstrating its robustness for water-air OWC 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.

Referee Report

2 major / 2 minor

Summary. The paper develops a mathematical channel model for time-varying refraction at the water-air interface in optical wireless communication and proposes a vision-based beam tracking algorithm that combines a convolutional neural network with bi-directional LSTM and an attention mechanism to extract spatio-temporal features from vision data. Simulations are presented to demonstrate that the algorithm outperforms classical methods in maintaining received signal strength and suppressing vision noise for underwater-to-low-altitude-platform links.

Significance. If the underlying channel model and synthetic noise accurately represent real oceanic conditions, the approach could support more reliable high-bandwidth backhaul for underwater applications. The work is simulation-driven with no reported experimental validation against measured wave-tank data or field trials, which constrains the strength of the robustness claims for deployed systems.

major comments (2)
  1. [Simulations / Performance Evaluation] The simulation results (described in the abstract and performance evaluation sections) rest exclusively on the authors' mathematical model of dynamic sea-surface refraction plus added synthetic vision noise. No experimental validation, wave-tank measurements, or field data are reported to confirm that the model's refraction statistics and beam-wander characteristics match real oceanic dynamics; this directly affects the load-bearing claim of robustness for water-air OWC systems.
  2. [Performance Evaluation] The abstract states that the algorithm 'outperforms classical methods in maintaining received signal strength and suppressing vision noise,' yet the manuscript provides no quantitative details on the classical baselines (e.g., Kalman filter parameters or beam-steering heuristics), error-bar reporting, or data-exclusion criteria used in the Monte-Carlo trials.
minor comments (2)
  1. [Channel Model] Notation for the sea-surface wave spectrum parameters and the refraction angle statistics should be defined explicitly in the channel-model section to allow reproducibility.
  2. [Figures] Figure captions for the simulation results should include the number of Monte-Carlo runs and the specific wave-height or wind-speed ranges used.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive feedback on our manuscript. We address the major comments point by point below, providing clarifications and committing to revisions that strengthen the presentation without overstating the simulation-based scope of the work.

read point-by-point responses

  1. Referee: [Simulations / Performance Evaluation] The simulation results (described in the abstract and performance evaluation sections) rest exclusively on the authors' mathematical model of dynamic sea-surface refraction plus added synthetic vision noise. No experimental validation, wave-tank measurements, or field data are reported to confirm that the model's refraction statistics and beam-wander characteristics match real oceanic dynamics; this directly affects the load-bearing claim of robustness for water-air OWC systems.

    Authors: We acknowledge that the evaluation relies on the proposed mathematical channel model for time-varying refraction combined with synthetic vision noise, as described in the manuscript. The model is derived from established physical principles of light propagation across the fluctuating water-air interface. We agree that the absence of wave-tank or field measurements limits direct confirmation of the model's statistical match to real oceanic conditions. In the revised manuscript we will add an explicit limitations subsection that states the simulation-based nature of the results, details the model's assumptions, and qualifies the robustness claims accordingly. We will also outline planned future experimental validation. revision: partial

  2. Referee: [Performance Evaluation] The abstract states that the algorithm 'outperforms classical methods in maintaining received signal strength and suppressing vision noise,' yet the manuscript provides no quantitative details on the classical baselines (e.g., Kalman filter parameters or beam-steering heuristics), error-bar reporting, or data-exclusion criteria used in the Monte-Carlo trials.

    Authors: We appreciate this observation on reproducibility. The revised manuscript will include a dedicated subsection in the performance evaluation that specifies the classical baseline implementations, including Kalman filter parameters, beam-steering heuristics, and any other methods compared. We will also add error bars to the simulation figures and report the Monte-Carlo trial count together with the data-exclusion criteria applied. revision: yes

standing simulated objections not resolved
  • The current study contains no experimental validation, wave-tank measurements, or field data to confirm the channel model's refraction statistics against real oceanic dynamics.

Circularity Check

0 steps flagged

No circularity: model and algorithm are independently specified; simulations are internal evaluation

full rationale

The paper first states a mathematical channel model for time-varying sea-surface refraction, then describes a CNN+Bi-LSTM+attention tracker that extracts spatio-temporal features from vision data generated under that model. The reported performance gains are obtained by running the tracker inside simulations driven by the same model and comparing it to classical methods. This constitutes standard simulation-based validation rather than any reduction of the algorithm's claimed behavior to a fitted parameter or self-citation by construction. No equation is shown to be tautological with its inputs, and no load-bearing uniqueness theorem or ansatz is imported from prior self-work. The derivation chain therefore remains self-contained against the paper's own stated assumptions and simulation environment.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on a domain-specific channel model for dynamic refraction and standard assumptions in neural network training for feature extraction; no free parameters or invented entities are explicitly listed in the abstract.

axioms (1)
  • domain assumption Dynamic refraction at the time-varying sea surface can be captured by a mathematical channel model that enables prediction of beam misalignment.
    Invoked to establish the foundation for the vision-based tracking algorithm.

pith-pipeline@v0.9.0 · 5746 in / 1193 out tokens · 34400 ms · 2026-05-21T22:10:29.130805+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

  1. [1]

    An overview of the internet of underwater things,

    M. C. Domingo, “An overview of the internet of underwater things,” Journal of Network and Computer Applications, vol. 35, pp. 1879–1890, Jul. 2012

    work page 2012

  2. [2]

    Real-time water-to-air communication system under dynamic water surface and strong background radiation,

    T. Lin, T. Wei, Q. Hu, C. Fu, N. Huang, X. Liu, L. Tang, L. Su, J. Luo, and C. Gong, “Real-time water-to-air communication system under dynamic water surface and strong background radiation,”IEEE Photonics Journal, vol. 16, pp. 1–10, May. 2024

    work page 2024

  3. [3]

    A survey of underwater optical wireless communications,

    Z. Zeng, S. Fu, H. Zhang, Y . Dong, and J. Cheng, “A survey of underwater optical wireless communications,”IEEE Communications Surveys & Tutorials, vol. 19, pp. 204–238, Oct. 2016

    work page 2016

  4. [4]

    Underwater and water-air optical wireless communication,

    L.-K. Chen, Y . Shao, and Y . Di, “Underwater and water-air optical wireless communication,”Journal of Lightwave Technology, vol. 40, pp. 1440–1452, Oct. 2022

    work page 2022

  5. [5]

    Underwater optical wireless communi- cation,

    H. Kaushal and G. Kaddoum, “Underwater optical wireless communi- cation,”IEEE Access, vol. 4, pp. 1518–1547, Apr. 2016

    work page 2016

  6. [6]

    A long distance underwater visible light communication system with single photon avalanche diode,

    C. Wang, H.-Y . Yu, and Y .-J. Zhu, “A long distance underwater visible light communication system with single photon avalanche diode,”IEEE Photonics Journal, vol. 8, pp. 1–11, Aug. 2016

    work page 2016

  7. [7]

    Experimental demonstration and practical application of a real-time full-duplex underwater wireless optical communication transceiver,

    C. Cai, Z. Du, W. Liao, Y . He, X. Ma, S. Qin, R. Wang, X. Wang, T. Zhang, G. Song, Q. Chen, Y . Zhang, Y . Gao, H. Wang, Z. Zhang, and J. Xu, “Experimental demonstration and practical application of a real-time full-duplex underwater wireless optical communication transceiver,”Journal of Lightwave Technology, vol. 42, pp. 8541–8554, Aug. 2024

    work page 2024

  8. [8]

    Recent progress of air/water cross-boundary communications for underwater sensor networks: A review,

    H. Luo, J. Wang, F. Bu, R. Ruby, K. Wu, and Z. Guo, “Recent progress of air/water cross-boundary communications for underwater sensor networks: A review,”IEEE Sensors Journal, vol. 22, pp. 8360– 8382, Mar. 2022

    work page 2022

  9. [9]

    Dual- hop mixed rf-uow communication system: A phy security analysis,

    E. Illi, F. El Bouanani, D. B. Da Costa, F. Ayoub, and U. S. Dias, “Dual- hop mixed rf-uow communication system: A phy security analysis,” IEEE Access, vol. 6, pp. 55345–55360, 2018

    work page 2018

  10. [10]

    A software-defined opto-acoustic network architecture for internet of underwater things,

    A. Celik, N. Saeed, B. Shihada, T. Y . Al-Naffouri, and M.-S. Alouini, “A software-defined opto-acoustic network architecture for internet of underwater things,”IEEE Communications Magazine, vol. 58, pp. 88– 94, Apr. 2020

    work page 2020

  11. [11]

    Multi-user scheduling and channel characteristics for water-to-air unmanned vehicular visible light communication,

    T. Wei, T. Lin, N. Huang, C. Fu, X. Liu, L. Su, L. Tang, Q. Hu, and C. Gong, “Multi-user scheduling and channel characteristics for water-to-air unmanned vehicular visible light communication,”IEEE Transactions on V ehicular Technology, vol. 73, pp. 18705–18718, Jul. 2024

    work page 2024

  12. [12]

    Mitigation of wave-induced packet loss for water-air optical wireless communication by a tracking system,

    Y . Di, Y . Shao, and L.-K. Chen, “Mitigation of wave-induced packet loss for water-air optical wireless communication by a tracking system,” in 2021 Optical Fiber Communications Conference and Exhibition (OFC), pp. 1–3, Jul. 2021

    work page 2021

  13. [13]

    Waving effect characterization for water-to-air optical wireless communication,

    T. Lin, C. Fu, T. Wei, N. Huang, X. Liu, L. Tang, L. Su, J. Luo, and C. Gong, “Waving effect characterization for water-to-air optical wireless communication,”Journal of Lightwave Technology, vol. 41, pp. 120–136, Oct. 2023

    work page 2023

  14. [14]

    Background noise resistant underwater wireless optical communica- tion using faraday atomic line laser and filter,

    J. Zhang, G. Gao, B. Wang, X. Guan, L. Yin, J. Chen, and B. Luo, “Background noise resistant underwater wireless optical communica- tion using faraday atomic line laser and filter,”Journal of Lightwave Technology, vol. 40, pp. 63–73, Jan. 2022

    work page 2022

  15. [15]

    Field demonstrations of wide-beam optical communications through water–air interface,

    X. Sun, M. Kong, O. A. Alkhazragi, K. Telegenov, M. Ouhssain, M. Sait, Y . Guo, B. H. Jones, J. S. Shamma, T. K. Ng, and B. S. Ooi, “Field demonstrations of wide-beam optical communications through water–air interface,”IEEE Access, vol. 8, pp. 160480–160489, Sep. 2020

    work page 2020

  16. [16]

    Preliminary characteri- zation of coverage for water-to-air visible light communication through wavy water surface,

    T. Lin, N. Huang, C. Gong, J. Luo, and Z. Xu, “Preliminary characteri- zation of coverage for water-to-air visible light communication through wavy water surface,”IEEE Photonics Journal, vol. 13, pp. 1–13, Jan. 2021. 9

    work page 2021

  17. [17]

    Real-time wave mitigation for water-air owc systems via beam tracking,

    Y . Di, Y . Shao, and L.-K. Chen, “Real-time wave mitigation for water-air owc systems via beam tracking,”IEEE Photonics Technology Letters, vol. 34, pp. 47–50, Dec. 2022

    work page 2022

  18. [18]

    Beam tracking aided by complexity-reduced MobileNetV2 for water-air OWC with waves,

    A. Xu, Y . Di, X. Yue, and L.-K. Chen, “Beam tracking aided by complexity-reduced MobileNetV2 for water-air OWC with waves,” IEEE Photonics Technology Letters, vol. 36, pp. 1469–1472, dec, 2024

    work page 2024

  19. [19]

    Bubble-wave-mitigation algorithm and transformer-based neural network demodulator for water- air optical camera communications,

    M. Li, P. Ling, S. Wen, X. Chen, and F. Wen, “Bubble-wave-mitigation algorithm and transformer-based neural network demodulator for water- air optical camera communications,”IEEE Photonics Journal, vol. 15, pp. 1–10, Aug. 2023

    work page 2023

  20. [20]

    Reinforcement-learning-enabled beam alignment for water-air direct optical wireless communications,

    J. Liu, T. Mao, D. He, Y . Yang, Z. Gao, D. Zheng, and J. Zhang, “Reinforcement-learning-enabled beam alignment for water-air direct optical wireless communications,” in2024 IEEE/CIC International Conference on Communications in China (ICCC), pp. 138–143, Sep. 2024

    work page 2024

  21. [21]

    Enabling reliable water–air direct optical wireless communication for uncrewed vehicular networks: A deep reinforcement learning approach,

    J. Wang, H. Luo, R. Ruby, J. Liu, S. Wang, and K. Wu, “Enabling reliable water–air direct optical wireless communication for uncrewed vehicular networks: A deep reinforcement learning approach,”IEEE Transactions on V ehicular Technology, vol. 73, pp. 11470–11486, Aug. 2024

    work page 2024

  22. [22]

    High-speed 645-nm VCSELs for low-scattering- loss gb/s underwater wireless optical communications,

    N. Almaymoni, O. Alkhazragi, W. H. Gunawan, G. Melinte, T. K. Ng, and B. S. Ooi, “High-speed 645-nm VCSELs for low-scattering- loss gb/s underwater wireless optical communications,”IEEE Photonics Technology Letters, vol. 36, pp. 377–380, Mar. 2024

    work page 2024

  23. [23]

    Three-dimensional dynamic sea surface modeling based on ocean wave spectrum,

    Z. Chang, F. Han, Z. Sun, Z. Gao, and L. Wang, “Three-dimensional dynamic sea surface modeling based on ocean wave spectrum,”Acta Oceanologica Sinica: English ver ., vol. 40, p. 11, Jun. 2021

    work page 2021

  24. [24]

    Research on underwater wireless optical communication channel model and its application,

    K. Sun, Y . Li, and Z. Han, “Research on underwater wireless optical communication channel model and its application,”Applied Sciences (Switzerland), vol. 14, Dec. 2023. Cited by: 2; All Open Access, Gold Open Access

    work page 2023

  25. [25]

    Underwater laser communication using an OFDM- modulated 520-nm laser diode,

    J. Xu, A. Lin, X. Yu, Y . Song, M. Kong, F. Qu, J. Han, W. Jia, and N. Deng, “Underwater laser communication using an OFDM- modulated 520-nm laser diode,”IEEE Photonics Technology Letters, vol. 28, pp. 2133–2136, Oct. 2016

    work page 2016

  26. [26]

    Improvement of the transmission bandwidth for indoor optical wireless communication systems using a diffused gaussian beam,

    D. Wu, Z. Ghassemlooy, H. L. Minh, S. Rajbhandari, and A. C. Boucou- valas, “Improvement of the transmission bandwidth for indoor optical wireless communication systems using a diffused gaussian beam,”IEEE Communications Letters, vol. 16, no. 8, pp. 1316–1319, 2012

    work page 2012

  27. [27]

    Study on led visible light communication channel model based on poisson stochastic network theory,

    H. Wu and Q. Fan, “Study on led visible light communication channel model based on poisson stochastic network theory,” in2020 Inter- national Conference on Wireless Communications and Smart Grid (ICWCSG), pp. 5–9, IEEE, Sep. 2020

    work page 2020

  28. [28]

    Absorption spectrum (380–700 nm) of pure water. ii. integrating cavity measurements,

    R. M. Pope and E. S. Fry, “Absorption spectrum (380–700 nm) of pure water. ii. integrating cavity measurements,”Applied optics, vol. 36, no. 33, pp. 8710–8723, 1997

    work page 1997

  29. [29]

    Fresnel equations,

    A. I. Lvovsky, “Fresnel equations,”Encyclopedia of Optical Engineer- ing, vol. 27, pp. 1–6, 2013

    work page 2013

  30. [30]

    Ray tracing on the cell processor,

    C. Benthin, I. Wald, M. Scherbaum, and H. Friedrich, “Ray tracing on the cell processor,” in2006 IEEE Symposium on Interactive Ray Tracing, pp. 15–23, Sep. 2006

    work page 2006

  31. [31]

    Imaging velocity fields analysis of space camera for dynamic circular scanning,

    T. Xu, X. Yang, S. Wang, J. Han, L. Chang, and W. Yue, “Imaging velocity fields analysis of space camera for dynamic circular scanning,” IEEE Access, vol. 8, pp. 191574–191585, Oct. 2020

    work page 2020

  32. [32]

    CNN, LSTM, and Bi-LSTM based self-attention model classification for user review sentiment analysis,

    K. S. Gill, V . Anand, R. Chauhan, A. Choudhary, and R. Gupta, “CNN, LSTM, and Bi-LSTM based self-attention model classification for user review sentiment analysis,” in2023 3rd International Conference on Smart Generation Computing, Communication and Networking (SMART GENCON), pp. 1–6, Dec. 2023

    work page 2023

  33. [33]

    Action recognition using visual attention with reinforcement learning(conference paper),

    H. Li, J. Chen, R. Hu, M. Yu, H. Chen, and Z. Xu, “Action recognition using visual attention with reinforcement learning(conference paper),” 2019 Action recognition using visual attention with reinforcement learn- ing (Conference Paper), pp. 365–376, Dec. 2018

    work page 2019