Statistical Characterization of Wind-Induced Beam Refraction in Water-to-Air Optical Channels
M. Nennouche, I. C. Ijeh, M. A. Khalighi · 2026
Read Statistical Characterization of Wind-Induced Beam Refraction in Water-to-Air Optical Channels by M. Nennouche, I. C. Ijeh, M. A. Khalighi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-22861.txt.
Introduction
Nennouche, Ijeh, and Khalighi model a vertical optical link from an underwater node to a drone. Wind bends the beam at the moving sea surface and shakes the drone itself. They fit wind-driven pointing errors with a Beta mixture. Then they derive closed-form formulas for the channel distribution and outage chance. Their account covers breaks from total internal reflection and a limited receiver field of view. It finds a reflection-driven outage floor that caps link reliability.
Catalog record
| CREATOR | M. Nennouche, I. C. Ijeh, M. A. Khalighi |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 22503 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.22861
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Statistical Characterization of Wind-Induced Beam Refraction in Water-to-Air Optical Channels M. Nennouche, I. C. Ijeh, M. A. Khalighi Abstract Direct water-to-air (W2A) optical communications experience strong beam refraction at the dynamic sea surface. This letter proposes a novel and tractable statistical channel model for a vertical W2A link between an underwater node and an unmanned aerial vehicle under varying wind speeds, modeling wind-induced pointing errors with a Beta mixture fitted via the Expectation–Maximization algorithm. By accounting for link interruptions due to total internal reflection (TIR) and receiver field-of-view limitations, we derive closed-form expressions for the channel distribution and link outage probability. Our analysis reveals a fundamental TIR-induced outage floor limiting link reliability and providing insight for robust W2A system design. I Introduction Underwater communications are attracting growing interest, driven by the rapid expansion of the internet of underwater things (IoUT) across diverse applications, ranging from industrial pipeline inspection to environmental coral reef monitoring [ 6 , 8 ] . These use-cases require data rates and bandwidths beyond the capabilities of conventional acoustic systems. Consequently, underwater wireless optical communications (UWOCs) have emerged as a compelling alternative, enabling high-speed data transmission over distances of several tens of meters [ 6 ] . In UWOC systems, retrieving data to the surface for remote processing remains a key challenge. The most common approach relies on surface platforms, such as buoys or autonomous surface vehicles [ 8 ] , introducing additional cost and environmental impact. As a result, direct water-to-air (W2A) communication has recently gained significant traction [ 8 ] . This approach, however, introduces increased channel complexity due to the dynamic nature of the sea surface. Wave-induced random refraction and reflection [ 8 , 12 ] make accurate modeling of the direct W2A channel highly challenging, yet such modeling is essential for ensuring reliable communication. Several recent studies have focused on modeling the direct W2A channel, primarily through Monte-Carlo (MC) simulations, developing realistic simulators that incorporate various sea surface models and underwater impairments, including absorption and scattering by suspended particles, as well as the effects of air bubbles generated by wave breaking [ 12 , 7 , 1 ] . To mitigate the high computational cost of MC -based simulations, recent works have proposed statistical channel models for both W2A [ 2 ] and air-to-water (A2W) [ 14 ] links. This work aligns with this ongoing research effort by providing a comprehensive modeling framework. We propose a model for optical beam refraction at the sea surface as a function of wind speed. In particular, we derive a novel closed-form and tractable expression to estimate the pointing error (PE) loss using a Beta mixture distribution, whose parameters are linked to wind speed. Furthermore, the link outage probability is evaluated, accounting for total internal reflection (TIR) . Our analytical results are validated through extensive MC simulations, enabling a detailed analysis of the outage performance. II Link Geometry and System Model We consider a vertical W2A link between a base station equipped with a laser diode (LD) at depth z_{w} , and an unmanned aerial vehicle (UAV) hovering at altitude z_{a} above the sea surface, equipped with a photodetector (PD) with field-of-view (FoV) \theta_{\mathrm{FoV}} . Perfect alignment between the transmitter (Tx) and receiver (Rx) is assumed, as illustrated in Fig. 1 . We consider a dynamic sea surface, with an incident angle \theta_{\mathrm{I}} defined between the surface normal \vec{n} and the LD beam, which is refracted to an angle \theta_{\mathrm{A}} with respect to the vertical axis, referred hereafter to as the angle-of-arrival (AoA) . Fig. 1: Illustration of the vertical water-to-air (W2A) optical wireless communication system model. We employ non-return-to-zero on-off keying (NRZ-OOK) modulation, where a high transmit power P_{\mathrm{Tx},1} is assigned to ‘1’ bits and a low level P_{\mathrm{Tx},0}=\xi P_{\mathrm{Tx},1} to ‘0’ bits, with \xi denoting the extinction ratio. At the Rx , a high-sensitivity silicon photo-multiplier (SiPM) is employed to enhance the transmission range, followed by a trans-impedance amplifier (TIA) , and a low-pass filter (LPF) to limit the Rx noise level. The overall channel loss, denoted by h , is formulated as: h=h_{\mathrm{L}}h_{\mathrm{P}}h_{\mathrm{A}}, (1) where h_{\mathrm{L}} represents the path loss, h_{\mathrm{P}} is the PE -induced loss, and h_{\mathrm{A}} accounts for the link interruption loss. Oceanic turbulence-induced scintillation is neglected in this study, as the large Rx aperture (i.e., the PD lens) significantly reduces irradiance fluctuations through aperture averaging [ 4 ] . Furthermore, given the short in-air propagation distance and the broadened beam footprint over the vertical underwater segment, the effects of atmospheric turbulence and underwater beam wandering are considered negligible compared to the dominant surface-induced refractions. III Water-to-Air Refraction Effect Modeling Assuming a static Tx and a UAV equipped with an attitude control system ensuring stable hovering [ 9 ] , the dominant source of PEs is the random motion of the sea surface. This results in fluctuations in the radial displacement r_{d} at the Rx plane. The resulting PE loss h_{\mathrm{P}} can be approximated as [ 4 ] : h_{\mathrm{P}}\approx A_{0}\exp\left(-2r_{d}^{2}/\omega_{\mathrm{Leq}}^{2}\right), (2) where A_{0}=\left(\mathrm{erf}(v)\right)^{2} represents the maximum fraction of received power (i.e., for r_{d}=0 ), with \mathrm{erf}(.) denoting the error function and v=\left(D_{r}\sqrt{\pi}\right)/\left(2\sqrt{2}~\omega_{L}\right) . Here, D_{r} is the PD lens diameter, \omega_{L} is the beam waist at the total propagation distance L=z_{w}+z_{a} , and \omega_{\mathrm{Leq}} denotes the equivalent beam width at the Rx : \omega_{\mathrm{Leq}}=\omega_{L}\sqrt{\sqrt{\pi}~\mathrm{erf}(v)/\left[2v\exp\left(-v^{2}\right)\right]}. (3) The probability density function (PDF) of h_{\mathrm{P}} is then: f_{h_{\mathrm{P}}}(h_{\mathrm{P}})=\frac{\omega_{\mathrm{Leq}}^{2}}{4h_{\mathrm{P}}\sqrt{-\frac{\omega_{\mathrm{Leq}}^{2}}{2}\ln\left(\frac{h_{\mathrm{P}}}{A_{0}}\right)}}f_{r_{d}}\left(\sqrt{-\frac{\omega_{\mathrm{Leq}}^{2}}{2}\ln\left(\frac{h_{\mathrm{P}}}{A_{0}}\right)}\right), (4) where f_{r_{d}}(\cdot) denotes the PDF of r_{d} , which depends on the AoA \theta_{\mathrm{A}} . Given that the beam deviation occurs at the surface, and assuming x_{\mathrm{Tx}}=y_{\mathrm{Tx}}=x_{\mathrm{Rx}}=y_{\mathrm{Rx}}=0 , the radial displacement can be expressed geometrically as r_{d}=z_{a}\tan\left(\theta_{\mathrm{A}}\right) . III-A AoA Statistical Modeling Before modeling the AoA , it is necessary to establish an accurate statistical model for the incident angle at the sea surface, \theta_{\mathrm{I}} , a topic that has recently received considerable attention. Prior studies have modeled the instantaneous surface motion using distributions such as Rayleigh [ 2 ] and, more recently, Weibull [ 5 ] . Owing to its model consideration of the spatio-temporal correlations of wave dynamics, and agreement with experimental sea measurements, the Weibull distribution is adopted in this work, with the corresponding PDF given by: f_{\theta_{\mathrm{I}}}(\theta_{\mathrm{I}})=\frac{k_{U}}{\lambda_{U}}\left(\frac{\theta_{\mathrm{I}}}{\lambda_{U}}\right)^{k_{U}-1}\exp\left[-\left(\frac{\theta_{\mathrm{I}}}{\lambda_{U}}\right)^{k_{U}}\right], (5) where k_{U}=1.7454+0.0071\,U and \lambda_{U}=13.6485+0.2406\,U are the PDF shape and scale parameters, respectively, valid for wind speeds U\in[6,15] m/s. Let \theta_{\mathrm{T}} denote the transmission angle relative to the surface normal \vec{n} , see Fig. 1 . The AoA can then be expressed as \theta_{\mathrm{A}}=\theta_{\mathrm{T}}-\theta_{\mathrm{I}} . The angle \theta_{\mathrm{T}} is obtained via Snell’s law as \theta_{\mathrm{T}}=\arcsin\left(\frac{n_{1}}{n_{2}}\sin(\theta_{\mathrm{I}})\right) , where n_{1}\approx 1.33 and n_{2}\approx 1 are the refractive indices of water and air, respectively [ 10 ] . Due to the highly non-linear nature of this geometric transformation, deriving a tractable, closed-form PDF for \theta_{\mathrm{A}} is mathematically prohibitive. Therefore, we opt for a robust statistical approximation. Using random numerical sampling for probability estimation, we evaluated several candidate distributions, including Rayleigh, Nakagami, and Weibull. The results, shown in Figs. 2a , 2b , and 2c , for U=6 , 10 , and 14 m/s, show that \theta_{\mathrm{A}} is best approximated by a Weibull distribution with shape and scale parameters denoted by k_{\mathrm{A}} and \lambda_{\mathrm{A}} , respectively. Furthermore, regression analysis indicates that, over the considered wind-speed range, the shape