REVIEW 3 major objections 5 minor 30 references
On the Vulnerability of Underwater Magnetic Induction Communication
T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read The paper claims that underwater magnetic induction links, usually considered secure, can be eavesdropped, and that the risk depends on the eavesdropper's position and orientation.
desk verdict Qualitative result holds—MI eavesdropping depends on geometry—but the seawater numbers are not validated because experiments were done in air and simulations used freshwater conductivity. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central object is the resonant coil pair and the angular dependence of magnetic flux through a coil. The paper uses the standard induced-voltage relation $V_{\mathrm{ind}} = 2\pi N A B \cos\theta$ with the flux-density expression $B = \mu_0 N I r^2 \cos\theta / (2(r^2+d^2)^{3/2})$ and an exponential skin-depth attenuation term for conducting media, so every result is ultimately a function of coil offset angle $\theta$ and distance $d$. Finite-element models of five eavesdropper configurations compute the magnetic flux density and the voltage induced in the legitimate receiver and the eavesdropper; those voltages feed the secrecy-capacity expression $\mathrm{SC} = \log_2(1+\mathrm{SNR}_{\mathrm{Rx}}) - \log_2(1+\mathrm{SNR}_E)$. The 90° null in $\cos\theta$ is what produces the paper's cleanest result: an orthogonally oriented eavesdropper receives no information, while an aligned one can.
What would settle it
Place the same three coils in a seawater tank at $\sigma \approx 4$ S/m, repeat Configurations 1–5, and compare the eavesdropper and receiver voltages; if the eavesdropper voltage at the claimed working distances falls below the receiver noise floor, or if the receiver-voltage changes disappear, the paper's quantitative vulnerability assessment does not transfer to real oceans.
Extended reading notes
Core claim
The paper's central claim is that underwater MI communication is genuinely vulnerable to passive eavesdropping, but the vulnerability is conditional, not absolute. An eavesdropper can extract information whenever its coil is close to the legitimate transmitter and oriented so that magnetic flux threads its coil; maximum received voltage occurs when the eavesdropper faces the transmitter, and reception drops to zero at a 90° orientation, where field lines are parallel to the coil plane. The finite-element simulations and the air-medium lab experiments both show these trends across five configurations (two position-based, three orientation-based). The paper further computes secrecy capacity from the received voltages and finds it falls as the eavesdropper approaches the transmitter or aligns with it, and rises when the eavesdropper is far or orthogonal. A second claimed finding is that the legitimate receiver's voltage changes in several of these configurations, which the paper presents as a possible basis for detecting malicious nodes.
Load-bearing premise
The quantitative results assume that air-medium lab tests and simulations at 0.01 S/m water conductivity faithfully represent real seawater, whose conductivity is roughly 4 S/m and whose eddy-current losses are much stronger.
Editorial extensions
If this is right
- A deployed MI link should treat the physical geometry of nearby coils as a security parameter: an aligned, close eavesdropper can read the traffic, so coil placement and orientation need active management.
- Legitimate receivers that can measure small induced-voltage changes have a built-in intrusion signal in several configurations, which could trigger countermeasures such as jamming or link re-orientation.
- Secrecy capacity predictions give operators a rule of thumb: keep the eavesdropper farther from the transmitter than the legitimate receiver, and avoid alignment between Tx and the eavesdropper's likely positions.
- The angular null at 90° means there are orientations in which an eavesdropper receives nothing; directional coil designs or deliberate misalignment could be used as a physical-layer hardening measure.
Reading between the lines
- In real seawater ($\sigma \approx 4$ S/m), the exponential attenuation term in the paper's Eq. (2) is far stronger than in the 0.01 S/m simulations, so the absolute voltages and secrecy-capacity values should be read as optimistic; the geometric null at 90° should persist, but the eavesdropper's usable range is likely shorter.
- The observed perturbation of the legitimate receiver's voltage could be developed into an active authentication scheme: the transmitter can vary its load or current in a known pattern, and the receiver can check whether the received-voltage signature matches, flagging any coil that disturbs it.
- A single-eavesdropper model may understate the threat: multiple malicious coils or a relay coil could detune or retune the field in ways that enlarge the eavesdropping region, and the paper does not model that case.
- The experiment was run at 100 kHz in air; at the lower frequencies typical of long-range underwater MI, skin depth grows, so the quantitative gap between air and seawater may narrow, which is a testable extension.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper analyzes passive eavesdropping attacks on underwater magnetic induction (MI) communication. It considers a three-node scenario with a legitimate transmitter, a legitimate receiver, and an eavesdropping node, and develops finite-element-method (FEM) simulations for five configurations that vary the eavesdropper's position and orientation relative to the legitimate nodes. The performance metrics are induced voltage, magnetic flux density, and secrecy capacity. Laboratory experiments performed in air are used to replicate the simulation configurations. The authors conclude that underwater MI communication is vulnerable to eavesdropping, that the vulnerability depends strongly on the eavesdropper's position and orientation, and that legitimate nodes may detect malicious activity through changes in the received voltage.
Significance. If the claims hold, this work makes a useful contribution by challenging the common assumption that MI communication is inherently secure or covert underwater, and by providing a systematic map of the geometric conditions under which eavesdropping is most effective. The FEM simulations and laboratory experiments are mutually independent checks against standard coil-coupling physics, and the paper does not fit free parameters to its conclusions. The qualitative agreement between simulation and experiment across five configurations is a genuine strength. However, the quantitative transfer of the results to real seawater is not established: the FEM model uses a freshwater conductivity and the experiments are all in air. The qualitative conclusion that an aligned nearby coil can intercept the signal is credible, since the angular nulls are geometric in origin, but the voltage magnitudes and secrecy-capacity values in the paper should not be read as validated for seawater without further work.
major comments (3)
- [§III-C, Figs. 13–14] The quantitative underwater claims, including the received-voltage magnitudes and the secrecy-capacity values in Fig. 11, are not supported for the intended seawater medium. The FEM model fixes water conductivity at 0.01 S/m (Table I), and the real-world experiments in Section III-C are conducted in air, as the authors acknowledge. At the 100 kHz operating frequency, seawater has a conductivity of about 4 S/m, giving a skin depth of roughly 0.8 m in Eq. (2), while the simulated conductivity gives a skin depth of about 16 m. Over the 4 ft to 7.5 ft distances used, the eddy-current exponential factor in Eq. (2) falls to roughly 0.2–0.05 in seawater, compared with about 0.9 in the low-conductivity simulation. This materially changes the reported voltage levels and hence the secrecy-capacity numbers. I request either a seawater-conductivity simulation (ideally a conductivity sweep from 0.01 to 5 S/m), or a saline-tank experimental validation, or an explicit and prominent statement that the quantitative results hold only for low-conductivity media. The conclusion should then be qualified accordingly. The qualitative dependence on position and orientation, including the angular null at orthogonal alignment, is likely to survive in seawater, but the quantitative vulnerability assessment is not yet validated.
- [§III-B6, Eq. (5), Fig. 11] The experimental section reports single measurements without error bars, repeated trials, or any quantitative agreement metric between the FEM and experimental results. The text states that the trends are 'almost similar' but does not quantify the deviations. Since these experiments are the only direct validation of the FEM model, the lack of uncertainty quantification makes it difficult to assess whether the observed discrepancies affect the paper's conclusions, especially the quantitative claims in Fig. 11. Please add repeated measurements with error bars for each configuration and a stated agreement criterion, such as normalized root-mean-square error or a tolerance band.
- [§III-B6, Eq. (5), Fig. 11] The secrecy-capacity evaluation does not specify the noise power sigma^2 used in Eq. (5), the load resistance R, or whether the same noise variance is assumed at the legitimate receiver and the eavesdropper. These parameters directly determine the secrecy-capacity values plotted in Fig. 11, which are a central quantitative output of the paper. The authors should state and justify these parameter choices, and preferably report sensitivity of the secrecy-capacity curves to the assumed noise power.
minor comments (5)
- [Eq. (4)] Equation (4) contains a typographical error: the magnetic permeability constant should be 4π × 10^-7 H/m, not 4π × 10^7 H/m.
- [§III-A] The phrase 'secret capacity' in Section III-A should be 'secrecy capacity' for consistency with the rest of the paper.
- [§III-B1] There is a typo in Section III-B1: 'eavsdropper' should be 'eavesdropper'.
- [§III-B4] In Configuration 4, the description labels d_Tx-E = 6 ft, but the geometry rotates the eavesdropper around the legitimate receiver, so the transmitter-to-eavesdropper distance changes as a function of θ_Rx-E. The text and figures should clarify this angle-dependent distance.
- [§III-C] The FEM implementation details, including the software used, mesh settings, and boundary conditions, are not reported. Adding these details would improve reproducibility, and for a security paper that proposes a new vulnerability assessment the lack of available code or data is a limitation worth acknowledging.
Circularity Check
No significant circularity: the vulnerability result is computed from standard MI physics and independently cross-checked by FEM and lab experiments; self-citations are background only.
full rationale
The paper's central claim—that an eavesdropper coil can intercept an MI link depending on its position and orientation—follows directly from the standard magnetoquasistatic model: Eq. (2) gives the magnetic field B, Eq. (3) gives the induced voltage V_ind = 2πNAB cos(θ), and Eq. (5) defines secrecy capacity in terms of measured voltages. The FEM simulations implement these same physical equations, and the lab experiments in air reproduce the predicted angular and distance trends. Crucially, no parameter is fitted to the simulation or experimental outputs to force the result; the angular null at θ = 90° is inherent to the cos(θ) factor in Eq. (3), and the distance dependence is inherent to the (r^2+d^2)^(-3/2) term in Eq. (2). The paper's self-citations ([2], [10], and related background) are used only to motivate MI communication and experimental practice, not as the load-bearing source of the vulnerability conclusion. The only significant weakness is external validity, not circularity: Section III-C admits the lab experiments were performed in air and Table I sets water conductivity to 0.01 S/m, far below seawater (~4 S/m), so the quantitative underwater extrapolation is not validated. That is a validity/conductivity limitation, not a logical circle. Hence no circular step is present; the derivation is self-contained with respect to its own equations and data.
Assumptions & free parameters
free parameters (2)
- water_conductivity =
0.01 S/m
- noise_power =
not specified
assumptions (3)
- domain assumption MI propagation in air and water is similar because relative permeability is close to 1, so air experiments transfer to underwater.
- standard math Magnetic field from a coil is described by quasi-static dipole and loop formulas such as Eqs. (1) to (4).
- domain assumption Eavesdropper and legitimate receiver experience identical noise power sigma^2.
Cite this review
Pith. "Pith review of On the Vulnerability of Underwater Magnetic Induction Communication." pith.science (2026). https://pith.science/paper/JPRYTQR6
@misc{pith2026250504249,
author = {Pith},
title = {Pith review of: On the Vulnerability of Underwater Magnetic Induction Communication},
year = {2026},
howpublished = {\url{https://pith.science/paper/JPRYTQR6}},
note = {Machine review of arXiv:2505.04249}
}
read the original abstract
Typical magnetic induction (MI) communication is commonly considered a secure underwater wireless communication (UWC) technology due to its non-audible and non-visible nature compared to acoustic and optical UWC technologies. However, vulnerabilities in communication systems inevitably exist and may lead to different types of attacks. In this paper, we investigate the eavesdropping attack in underwater MI communication to quantitatively measure the system's vulnerability under this attack. We consider different potential eavesdropping configuration setups based on the positions and orientations of the eavesdropper node to investigate how they impact the received voltage and secrecy at the legitimate receiver node. To this end, we develop finite-element-method-based simulation models for each configuration in an underwater environment and evaluate the received voltage and the secrecy capacity against different system parameters such as magnetic flux, magnetic flux density, distance, and orientation sensitivity. Furthermore, we construct an experimental setup within a laboratory environment to replicate the simulation experiments. Both simulation and lab experimental confirm the susceptibility of underwater MI communication to eavesdropping attacks. However, this vulnerability is highly dependent on the position and orientation of the coil between the eavesdropper and the legitimate transmitter. On the positive side, we also observe a unique behavior in the received coil reception that might be used to detect malicious node activities in the vicinity, which might lead to a potential security mechanism against eavesdropping attacks.
Figures
Figures from the paper (10 more)
Reference graph
Works this paper leans on
-
[1]
Y . Li, S. Wang, C. Jin, Y . Zhang, and T. Jiang, “A survey of under- water magnetic induction communications: Fundamental issues, recent advances, and challenges,”IEEE Communications Surveys & Tutorials, 2019. (a) (b) (c) Fig. 14: Experimental setup based received voltage in legiti- mate Rx node and eavesdropper node vs. different eavesdrop- per node ori...
work page 2019
-
[2]
M. Muzzammil, N. Ahmed, G. Qiao, I. Ullah, and L. Wan, “Fun- damentals and advancements of magnetic field communication for underwater wireless sensor networks,”IEEE Transactions on Antennas and Propagation, 2020
work page 2020
-
[3]
Recent advances and future directions on underwater wireless commu- nications,
M. F. Ali, D. N. K. Jayakody, Y . A. Chursin, S. Affes, and S. Dmitry, “Recent advances and future directions on underwater wireless commu- nications,”Archives of Computational Methods in Engineering, vol. 27, pp. 1379–1412, 2020
work page 2020
-
[4]
Towards Sustainable Internet of Underwater Things: UAV-aided Energy Efficient Wake-up Solutions
M. Muzzammil, N. Kouzayha, N. Saeed, and T. Y . Al-Naffouri, “To- wards sustainable internet of underwater things: UA V-aided energy efficient wake-up solutions,”arXiv preprint arXiv:2208.12065, 2022
work page Pith review arXiv 2022
-
[5]
A survey of underwater wireless communication technologies,
C. M. Gussen, P. S. Diniz, M. L. Campos, W. A. Martins, F. M. Costa, and J. N. Gois, “A survey of underwater wireless communication technologies,”J. Commun. Inf. Sys, vol. 31, no. 1, pp. 242–255, 2016
2016
-
[6]
W. Aman, S. Al-Kuwari, M. Muzzammil, M. M. U. Rahman, and A. Ku- mar, “Security of underwater and air–water wireless communication: State-of-the-art, challenges and outlook,”Ad Hoc Networks, vol. 142, p. 103114, 2023
work page 2023
-
[7]
Fine doppler scale estimations for an underwater acoustic CP-OFDM system,
L. Wan, H. Jia, F. Zhou, M. Muzzammil, T. Li, and Y . Huang, “Fine doppler scale estimations for an underwater acoustic CP-OFDM system,” Signal Processing, vol. 170, p. 107439, 2020
work page 2020
-
[8]
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, no. 1, pp. 204–238, 2017
work page 2017
Show all 30 references
-
[9]
Realizing underwater communi- cation through magnetic induction,
I. F. Akyildiz, P. Wang, and Z. Sun, “Realizing underwater communi- cation through magnetic induction,”IEEE Communications Magazine, vol. 53, no. 11, pp. 42–48, 2015
2015
-
[10]
Directivity pattern of different coil structures for magneto-coupled communication systems,
M. Muzzammil, Z. Babar, N. Ahmed, G. Qiao, and S. Liu, “Directivity pattern of different coil structures for magneto-coupled communication systems,” inOCEANS 2019-Marseille. IEEE, 2019, pp. 1–4
2019
-
[11]
Maximizing secrecy rate of an orthogonal frequency divi- sion multiplexing-based multihop underwater acoustic sensor network,
W. Aman, M. M. Rahman, Z. Haider, J. Qadir, M. W. Nawaz, and G. A. Sidhu, “Maximizing secrecy rate of an orthogonal frequency divi- sion multiplexing-based multihop underwater acoustic sensor network,” Transactions on Emerging Telecommunications Technologies, vol. 31, no. 11, ...
2020
-
[12]
On modeling eavesdropping attacks in underwater acoustic sensor networks,
Q. Wang, H.-N. Dai, X. Li, H. Wang, and H. Xiao, “On modeling eavesdropping attacks in underwater acoustic sensor networks,”Sensors, vol. 16, no. 5, p. 721, 2016
2016
-
[13]
On securing underwater acoustic networks: A survey,
S. Jiang, “On securing underwater acoustic networks: A survey,”IEEE Communications Surveys & Tutorials, vol. 21, no. 1, pp. 729–752, 2018
2018
-
[14]
Analysis of security attacks and taxonomy in under- water wireless sensor networks,
I. Ahmad, T. Rahman, A. Zeb, I. Khan, I. Ullah, H. Hamam, and O. Cheikhrouhou, “Analysis of security attacks and taxonomy in under- water wireless sensor networks,”Wireless Communications and Mobile Computing, vol. 2021, no. 1, p. 1444024, 2021
2021
-
[15]
Security weaknesses of underwater wireless optical communication,
M. Kong, J. Wang, Y . Chen, T. Ali, R. Sarwar, Y . Qiu, S. Wang, J. Han, and J. Xu, “Security weaknesses of underwater wireless optical communication,”Optics Express, vol. 25, no. 18, pp. 21 509–21 518, 2017
2017
-
[16]
The security in optical wireless communication: A survey,
X. Zhang, G. Klevering, X. Lei, Y . Hu, L. Xiao, and G.-H. Tu, “The security in optical wireless communication: A survey,”ACM Computing Surveys, vol. 55, no. 14s, pp. 1–36, 2023
2023
-
[17]
Impact of scattering on secrecy outage probability of underwater optical wireless links,
R. Boluda-Ruiz, P. Salcedo-Serrano, B. Castillo-V ´azquez, A. Garc ´ıa- Zambrana, and J. M. Garrido-Balsells, “Impact of scattering on secrecy outage probability of underwater optical wireless links,”IEEE Journal of Oceanic Engineering, 2023
2023
-
[18]
Security of the near field communication protocol: an overview,
D. Nelson, M. Qiao, and A. Carpenter, “Security of the near field communication protocol: an overview,”Journal of Computing Sciences in Colleges, vol. 29, no. 2, pp. 94–104, 2013
2013
-
[19]
Classification and analysis of security attacks in near field communication,
M. RAHMAN and H. ELMILIGI, “Classification and analysis of security attacks in near field communication,”International Journal of Business and Cyber Security, vol. 1, no. 2, 2017
2017
-
[20]
Magnetic resonant beamforming for secured wireless power transfer,
H. Sun, H. Lin, F. Zhu, and F. Gao, “Magnetic resonant beamforming for secured wireless power transfer,”IEEE Signal Processing Letters, vol. 24, no. 8, pp. 1173–1177, 2017
2017
-
[21]
Towards defending eavesdropping on NFC,
A. A. Al Islam, T. Chakraborty, T. A. Khan, M. Zoraf, and C. S. Hyder, “Towards defending eavesdropping on NFC,”Journal of Network and Computer Applications, vol. 100, pp. 11–23, 2017
2017
-
[22]
Near field communication (NFC) technology security vulnerabilities and countermeasures,
M. M. Singh, K. Adzman, and R. Hassan, “Near field communication (NFC) technology security vulnerabilities and countermeasures,”Inter- national Journal of Engineering & Technology, vol. 7, no. 4.31, pp. 298–305, 2018
2018
-
[23]
Common attacks on near field communication technology,
A. H. Alrobaish, W. F. Al-mutairi, H. A. Alsuqayhi, and D. M. Ibrahim, “Common attacks on near field communication technology,” in2022 2nd International Conference on Computing and Information Technology (ICCIT). IEEE, 2022, pp. 110–114
2022
-
[24]
Secure capacity analysis for magnetic inductive coupling-based swipt system,
S. Han, H.-J. Kim, J. Lee, and J.-W. Choi, “Secure capacity analysis for magnetic inductive coupling-based swipt system,”IEEE Access, vol. 6, pp. 49 182–49 191, 2018
2018
-
[25]
Characterization of a 3d underwater magneto-inductive transmitter coil array,
S. P. Ravindran, J.-F. Bousquet, and N. Gaoding, “Characterization of a 3d underwater magneto-inductive transmitter coil array,” inOCEANS 2018 MTS/IEEE Charleston. IEEE, 2018, pp. 1–6
2018
-
[26]
Underwater communication employing high- sensitive magnetic field detectors,
M. Hott and P. A. Hoeher, “Underwater communication employing high- sensitive magnetic field detectors,”IEEE Access, vol. 8, pp. 177 385– 177 394, 2020
2020
-
[27]
Design and evaluation of low-cost and energy-efficient magneto-inductive sensor nodes for wireless sensor networks,
N. Ahmed, A. Radchenko, D. Pommerenke, and Y . R. Zheng, “Design and evaluation of low-cost and energy-efficient magneto-inductive sensor nodes for wireless sensor networks,”IEEE Systems Journal, no. 99, pp. 1–10, 2018
2018
-
[28]
Power equations and capacity performance of magnetic induction communication systems,
J. I. Agbinya and M. Masihpour, “Power equations and capacity performance of magnetic induction communication systems,”Wireless Personal Communications, vol. 64, pp. 831–845, 2012
2012
-
[29]
Magnetic induction for underwater wireless commu- nication networks,
M. C. Domingo, “Magnetic induction for underwater wireless commu- nication networks,”IEEE Transactions on Antennas and Propagation, vol. 60, no. 6, pp. 2929–2939, 2012
2012
-
[30]
Dy- namic magnetic induction wireless communications for autonomous- underwater-vehicle-assisted underwater IoT,
D. Wei, L. Yan, C. Huang, J. Wang, J. Chen, M. Pan, and Y . Fang, “Dy- namic magnetic induction wireless communications for autonomous- underwater-vehicle-assisted underwater IoT,”IEEE Internet of Things Journal, vol. 7, no. 10, pp. 9834–9845, 2020
2020
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.