About Us

The argument in favor of using filler text goes something like this: If you use real content in the Consulting Process, anytime you reach a review

Contact Info

  • Chicago 12, Melborne City, USA
  • (111) 111-111-1111
  • Example@gmail.com
  • Week Days: 09.00 to 18.00 Sunday: Closed

Accepted Papers and Proceedings | UComms

  • Home
  • Accepted Papers and Proceedings
// Authors

Accepted

Papers and Proceedings

Top

Technical Session: “Channel Measurements, Modelling and Simulation”

Authors:  J. Johnson, S. Singh, P. Traykovski, T. Duda, M. Porter, J. Peterson

Abstract: Understanding sound behavior in nearshore constrained environments such as ports and harbors renders unique challenges. Improving sound models in these channels offers many advantages, such as increasing the effectiveness of model-based design of underwater acoustic systems that enable harbor security. Our ongoing efforts focus on acoustically characterizing these complicated nearshore channels, analyzing acoustic communication system performance, and refining a 3D beam tracing propagation model for simulating modem transmission. Measurements collected in an industrialized estuarine harbor demonstrate sound propagation effects not only from the boundary features (i.e. bathymetry, surrounding human-made obstacles) but also from the environment as a function of varying bottom type and coastal water column SSP effects. Field observations are compared with 3D sound model simulations to identify feature and boundary reflection properties that contribute or dominate measured echo structure. A verified comprehensive modeling ability for these dynamic, and acoustically distinct, constrained environments is ideal for executing successful underwater operations such as utilizing acoustic modem networks.

Authors: F. Socheleau, P. van Walree

Abstract: This paper presents an optimal transport-based (OT) approach for augmenting underwater acoustic channel datasets through the interpolation of time-varying impulse responses. While existing datasets often suffer from limited diversity due to the high cost of real-world measurements, OT provides a geometrically principled framework to generate physically plausible synthetic TVIRs. The proposed method focuses on spatial interpolation between TVIRs acquired simultaneously at different sensor locations, serving as a proof of concept for OT-based dataset augmentation. Experimental results validate the effectiveness of the approach in enhancing dataset variability and improving the robustness of underwater communication system evaluations.

Authors: C. Pelekanakis, H. Dol, T. Melodia

Abstract: This paper presents a physics-based wideband underwater acoustic (UWA) channel simulator for generating time-varying channel impulse responses (CIRs). The simulator uses eigenray computation together with measured directional sea-surface wave spectra, allowing realistic surface motion to induce time-varying micro-multipath. The simulator is validated against real channel-probe measurements collected in a littoral environment in the 10-15 kHz band. The statistical analysis shows very good agreement in the dominant multipath structure, power delay profile (PDP) shape, and root-mean-square (RMS) delay spread. The simulator also reproduces the coherence time with reasonable accuracy. Remaining discrepancies in Doppler behavior and temporal coherence indicate that improved modeling of rough-surface reflection loss and bubble-mediated extinction is needed for higher-fidelity prediction.

Authors: G. Benson, T. Bell, S. Brownlee, T. Leonard, J. Imran, J. Davies, C. Weaver

Abstract: Underwater acoustic channels exhibit strong time and spatial variability, complicating estimation. This study presents fixed geometry measurements from the BUTEC range and analyses channel variability using compact scalar metrics. This study introduces CSiNR as a real-time receiver relevant channel complexity metric, derived from channel probe measurements, capturing multipath structure effects on coherent receiver operation. Results show channel characteristics cluster by geometry, enabling the reduced dimension representations that support efficient identification of representative channels for receiver, waveform and network testing.

Authors: E. Voisin, Z. Wang

Abstract: Standardized underwater acoustic communication protocols such as JANUS are increasingly used alongside mobile underwater platforms. This combination creates a need for protocol testing under channels where multipath geometry, per-path Doppler, and propagation delay all evolve during a single packet. OpenCREST (Open Channel Real-time Emulation Streaming Tool) addresses this need by simulating geometry-driven channels in real time on a commodity desktop PC while keeping the modem’s complete analog chain in the loop. The channel model uses the method of images in an isovelocity scene with a per-tap Farrow interpolator, so per-path Doppler emerges naturally from time-varying multipath delays. Three experiments assessed OpenCREST’s performance: two-way ranging maintained sub-meter standard deviation across four channel configurations; a recorded channel (BCH1) and its geometric twin delineated where each channel mode applies; and, in an approaching-platform scenario, JANUS packets reproduced range- and sea-state-dependent packet error rates consistent with evolving multipath. OpenCREST is released as an open-source repository for independent validation and community extension.

Authors: D. Egnor, M. Stojanovic, J. Preisig

Abstract: Underwater acoustic paths that reflect off a moving surface often result in Doppler spread. Surface waves are characterized by wave variance spectra whose parameters are available from oceanographic models. This paper uses analytical expressions to demonstrate a method for calculating Doppler spread from these parameters. The results are compared to measurements from a two-week field experiment, showing good agreement and following general trends over time. This simple, computationally expedient, and effective method for predicting the Doppler spread requires only a single propagation model run and no surface wave height realizations to be generated.

Top

Technical Session: “Game-changers, Paradigm Shifters, the Breaking Wave in Underwater Communications”

Authors: H. Hu

Abstract: Underwater optical wireless communication (UWOC) offers high data rate and low-latency links for short- and medium-range maritime networks, but practical deployment is constrained by pointing, acquisition, and tracking; water-dependent absorption and scattering; mobility of autonomous underwater vehicles; and the difficulty of maintaining narrow optical beams with mechanically steered optics. This paper proposes an integrated optical phased array (OPA) based UWOC architecture in which a chip-scale OPA provides electronically fast-steerable beams in the blue-green optical window. The core idea is to replace wide beams or fixed collimators or mechanical gimbals with a programmable optical antenna capable of rapid beam sweeping, closed-loop tracking, adaptive beamwidth control, angular filtering at the receiver, and multi-node spatial scheduling. The proposed concept is positioned as an enabling technology for agile UWOC links among AUVs, sensor nodes, docking stations, and hybrid acoustic-optical underwater networks.

Authors: J. Mulholland, I. Smolyaninov

Abstract: An overview and demonstrations of surface-plasmon electromagnetic waves excited at radio frequencies in various underwater environments is provided. The paper introduces the theory, modelling and observed results from various scenarios. This data is then compared against the expected bulk propagation loss through the water, accounting for ohmic absorption and spherical spreading. Three experimental cases are considered, to elicit the surface propagation effect, observing communication being achieved at distances many times the theoretical bulk propagation range predicted for each, supporting claims related to the surface plasmonic-based propagation coupling, and indicating areas where such capabilities may be advantageous over alternative underwater wireless communications techniques, as well as proposing future experiments to help further the understanding of the observed phenomena.

Authors: F. Reiser, F.O. Sbrisny, P. Durdaut, L.A. Sander, H. Francke, R. Kurberg, R. Adelung, D. Hugenbusch

Abstract: Conventional piezoceramic underwater sound transmitters are relatively heavy when generating low-frequency signals, limiting their deployment on autonomous and small marine platforms. This is a critical limitation, as low frequencies enable greater transmission ranges, significantly enhancing the effectiveness of underwater communication and acoustic surveillance systems. The NEPTUNE project (NEw ProjecTors for unmanned UNderwater systEms) develops an innovative underwater sound transmission method using Aerographene nanotechnology. This electrically conductive, sponge-like structure consists of a three-dimensional graphene nanotube network comprising up to 99.9% air or gas [1]. Rapid electrical heating induces explosive volume expansion, generating acoustic pressure waves. These waves drive a membrane that transmits sound into water. A functional prototype with embedded Aerographene has successfully demonstrated the working principle. Initial testing included underwater signal measurements and characterization of pressure characteristics inside the enclosure. While detected signals remain insufficient for commercial deployment, the Proof-of-Concept confirms technical feasibility.

Authors: A. Silva, F. Zabel, J. Gomes, R. Viegas

Abstract: Underwater acoustic networks are constrained by low bandwidth, long propagation delays, multipath, and spatially distributed interference, namely due to neighboring nodes. Directional transmission can improve the signal-to-interference-plus-noise ratio and increase spatial reuse, but conventional phased arrays are often too large and complex for compact underwater sensor nodes. This paper presents a design and feasibility study of a single spiral acoustic projector that combines calibrated circular and spiral radiation modes to synthesize an electronically steerable directional pattern. The study focuses on the transducer excitation concept, radiation-field control, and its potential integration within a networking architecture. Finite-element simulations show that the maximum response direction can be electronically rotated while preserving coherent phase along the selected bearing. Finally, a conceptual protocol framework is outlined in which circular/spiral signaling supports azimuth estimation and directional transmission carries data, indicating a path toward medium access control schemes that exploit spatial reuse in small-scale underwater acoustic networks.

Authors: D. Spinosa, F. Marin, J. Lazzarin, A. Montanari, F. Campagnaro, M. Zorzi

Abstract: The variety of technological applications that can benefit underwater missions—such as diver coordination and surveillance of critical areas with Autonomous Underwater Vehicles (AUVs)—calls for the development of a flexible underwater communication platform that can be customized according to the user needs. Building on the architecture of the low-cost SuM software-defined acoustic modem [1] developed for research purposes, we introduce SuM2, a new and more sophisticated acoustic device designed for operational conditions. SuM2, while maintaining interoperability with JANUS digital communication, enhances both the hardware and software architecture of the original SuM modem by introducing higher transmission power, an improved analog front-end, support for Single-Sideband (SSB) analog voice and advanced digital transmission features including multiplexed data streams and optional CSMA-like medium access. Sea trials demonstrate interoperability with commercial communication devices, reliable communication at ranges up to 3.3 km, and bit rates up to 19.2 kbps at short range. The results confirm that SuM2 achieves a significantly extended communication range compared to its predecessor while retaining full software-defined flexibility and compatibility with commercial SSB devices and NATO standards. The presented architecture establishes SuM2 as a versatile and affordable acoustic modem suitable for both research and operational deployments.

Top

Technical Session: “Integrated Sensing and Communications”

Authors: L. Schattenhofer, C. von Brandis, T. Kwasnitschka, K. Heger, E. González, C. Renner, J. Karstens, O. Landsiedel

Abstract: Acoustic ranging is the established approach for localizing seafloor sensor nodes and generally achieves high precision. However, performance evaluation typically relies on the same acoustic ranges and GNSS positions used during localization, making error estimates self-referential. We address this circularity by introducing a photogrammetric benchmark, georeferenced by a ship-based USBL system, that is constructed independently of the acoustic ranges and GNSS positions under evaluation. For a five-node acoustic sensor network deployed at 60 m depth offshore Santorini, Greece, we build a georeferenced photo mosaic of the deployment area using a ship-towed camera sled and extract relative node positions at centimeter-scale precision. We then compare three established localization algorithms, a centralized least-squares solver and two distributed approaches based on the extended Kalman filter and belief propagation, against this reference. All algorithms achieve relative accuracy below 2 m. The centralized least-squares approach outperforms the distributed approaches by more than 1 m. This demonstrates that the benchmark is sensitive enough to evaluate the performance of different algorithms, offering a path toward non-circular evaluation of underwater localization systems.

Authors: R. Viegas, F. Zabel, A. Silva, J. Gomes

Abstract: The use of underwater spiral acoustic fields is a promising approach for underwater unmanned vehicle (UUV) localization. This technique relies on the transmission of both circular and spiral acoustic fields, enabling the estimation of the departure bearing angle through their phase difference. However, the accuracy of this estimation is strongly influenced by the characteristics of the underwater acoustic channel and the multiplexing strategy employed. This work introduces a novel multiplexing technique for circular and spiral acoustic fields based on low cross-correlation signals. The proposed method is experimentally validated in a pool using a beacon setup consisting of a static spiral acoustic source and a single mobile hydrophone, whose position is acoustically estimated. Ground-truth localization is obtained from top-view video recordings of the pool. Three multiplexing strategies are evaluated and compared: time-frequency division multiplexing (TFDM), code division multiplexing (CDM), and X-division multiplexing (XDM). In general, experimental results show good agreement between estimated and ground-truth azimuth values. However, azimuth estimation accuracy and dispersion differ significantly: TFDM and XDM exhibit lower accuracy and higher variance compared to CDM approaches, while both CDM variants demonstrate consistent and superior angular performance. Overall, the results indicate that CDM-based multiplexing of circular and spiral fields is a promising candidate to enable simultaneous data transmission and single-receiver localization using a spiral source.

Authors: P. Oppermann, C. Renner

Abstract: Localization of underwater objects is required for numerous applications, including environmental monitoring, ropeless fishing, and equipment tracking. Ease of deployment and small, low-cost tracers running for years on a battery are incompatible with state-of-the-art solutions, especially in highly challenging multipath environments. We present a passive, soft-information-based localization method leveraging receiver mobility to estimate the 2D-location of a beacon signal with Angle of Arrival observations. The method is shown to be more robust than approaches that make hard decisions on time of arrival observations early on. Furthermore, we present the results in a real-world harbor environment, showing that median localization accuracy below 5 m can be achieved.

Top

Technical Session: “Interoperability, Networks and Adaptive Solutions for the Underwater Domain”

Authors: E. Altamiranda, V. Eide, N. Judell

Abstract: Underwater acoustic communication standards have progressed considerably at the physical and packet-signaling layers, yet consistent system-level behavior across independent implementations is not guaranteed by these standards alone. Building on STANAG-4748 and SWiGacoustic Level 1, this paper presents a reference design for dual-channel underwater acoustic networks that targets system-level interoperability rather than only signal-level compatibility. The design separates a low-rate control plane from a high-rate data plane, defines explicit medium-access and coordination mechanisms, and realizes networking directly at the link layer through frame tunneling. High data-rate communication is scheduled by the control plane and enforced through MAC behavior, enabling predictable multi-node interaction. The design is informed by implementation experience and prior feasibility tests of the Dual Channel Acoustic Protocol (DCAP). Observations and insights are reported rather than a full system-level performance evaluation and discuss how the proposed architecture can be combined with existing and evolving standards.

Authors: B. Parrein, L. Toutain, D. Ochkas, A. Montanari, A. Pelov

Abstract: The Internet Protocol (IP) is not common in the Internet of Underwater Things (IoUT) and, specifically, in underwater acoustic sensor network (UASN). This is mainly due to the extreme constraints on bit rate, energy, and latency. However, IP remains the most widely adopted protocol for ensuring layer 3 (L3) interoperability. In this paper, we promote the use of Static Context Header Compression (SCHC) to enable IPv6 for underwater applications as an alternative to JANUS standard. We demonstrate the feasibility of this approach, enabling end-to-end IPv6 connectivity from the shore to underwater end-devices and between end-devices even with small data payload as low as 2 Bytes. This work offers a perspective for enabling any protocols from the Internet Engineering Task Force (IETF) to IoUT.

Authors: F. Mahieu, N. Morozs, P. Mitchell, T. Tozer

Abstract: Efficient end-to-end data delivery in multi-hop relay networks requires selecting routes that maximize throughput while minimizing delay. In underwater acoustic networks (UANs), this challenge is exacerbated by the restrictive communication environment. Route selection must balance two opposite approaches: increasing the number of relay hops, which accumulates delay due to repeated transmission overhead, or reducing hop count by skipping intermediate nodes, which increases transmission distances and packet error rates (PER) due to signal degradation, leading to higher retransmission rates. To address this, we introduce the Q-learned Uneven Linear Routing (QULR) protocol, which uses a Q-learning algorithm to dynamically balance these trade-offs and maintain an optimal route tailored for linear network topologies. This is particularly relevant for subsea pipeline and cable monitoring applications, where a chain of relay nodes maintains connectivity with end-devices over large distances. Simulations demonstrate that QULR consistently outperforms simpler approaches in throughput and retransmission costs, resulting in lower energy requirements, while remaining adaptable to evolving network conditions.

Authors: F. Pacheco, B. Ferreira, N. Cruz

Abstract: Hybrid maritime networks must provide both surface communication and underwater localization despite deployment geometries that impose conflicting requirements on coverage, connectivity, and measurement diversity. This work evaluates grid, hexagonal, Poisson point process, and centroidal Voronoi surface-buoy deployments using physics-based RF and UA models. The results show that RF and UA coverage are primarily governed by deployment density, whereas connectivity and localization depend on spatial geometry. A germ–grain and percolation-inspired surrogate model is introduced to capture these deployment trends using a compact set of physically interpretable parameters. The model achieved validation R^2 values of 0.95–0.99 for RF and UA coverage, 0.56–0.97 for RF connectivity, and 0.87–0.96 for UA localization. The resulting framework provides a physically grounded, low-cost approximation for screening candidate deployments and enabling inverse deployment design before computationally intensive simulations.

Top

  Technical Session: “Machine Learning in Underwater Communications and Networks”

Authors: T. Paviet-Salomon, P.J. Bouvet, A. Pottier, B. Tomasi

Abstract: Underwater platforms, such as drones and stationary systems, can be equipped with multiple types of front-ends (e.g., acoustic, optical, or EM) as part of an underwater wireless communication system. Two mode-selection algorithms that automatically select the communication mode (either acoustic or optical) based on the distance between transmitter and receiver are proposed in this paper. One algorithm is based on a simple heuristic rule, the other is based on the gradient bandit learning algorithm. The communication performance in terms of the total amount of data transferred across a linear trajectory, is numerically analyzed. The proposed algorithms are both suitable to be implemented onboard of resource-constrained battery-powered platforms.

Authors: M. R. Tanhatalab, P. Casari

Abstract: Mobility management in underwater acoustic networks is challenging because of long propagation delays, limited bandwidth, fluctuating underwater noise, and energy constraints. Conventional threshold-based handover methods perform poorly under these dynamic conditions. This paper proposes a propagation delay-aware reinforcement learning (RL) handover framework for autonomous underwater vehicles (AUVs) served by surface gateways (BS). Our method uses spatiotemporal noise predictions from the marine noise analysis tool (MNAT) to support adaptive BS selection under predicted underwater acoustic noise conditions. Predictive noise information is integrated into the RL state space, reward formulation, and policy optimization process. Simulation results show that an integral reinforcement learning (IRL) approach provides the strongest robustness, attaining the highest average composite link quality index (CLQI) and throughput, while maintaining stable handover behavior under changing noise conditions.

Authors: G. A. Leaku, P. Casari

Abstract: We propose an acoustic bandwidth-aware design for a deep neural network (DNN) for image classification that splits inference between an energy-constrained device and a second, more powerful system, in order to save energy on the former. Our method integrates quantization and sparsification of transmitted feature maps to scale communication efforts down with minimal accuracy reduction. We evaluate our design using the CIFAR-100 dataset and the Aqua20 underwater image dataset. Our results show that optimal tradeoffs enabled by different split points yield over 80% reduction in transmission time with a <4% accuracy drop by applying the proposed optimizations.

Authors: S. Yildirim, E. Chinellato, R. Petroccia

Abstract: This work investigates adaptive communications in underwater acoustic channels and proposes a neural network (NN) model based on a bank of configuration-specific networks, with one network dedicated to each signal configuration. Each network consists of eight fully connected layers and predicts the symbol error rate (SER) using a set of input features, including pre- and post-equalization signal to noise ratios (SNR), peak-to-average power ratio, channel delay and Doppler spreads, and Doppler shift normalized by the carrier frequency. We evaluate the proposed approach against a binary decision tree (DT) method in terms of SER prediction, generalization capability, and achieved throughput. The sensitivity of the proposed NN to shifts in normalization statistics is evaluated. The NN framework closely tracks the measured throughput, achieving an order of magnitude improvement in SER prediction accuracy in terms of mean squared error (MSE), whereas the DT overestimates the throughput by a factor of three at low SNRs.

 

Authors: D. A. Cuji, A. Singer, M. Stojanovic

Abstract: Transmit beamforming for underwater acoustic communications is challenging because channel state information must be obtained through feedback and may be noisy or outdated. In this paper, we investigate an angle-based beamforming strategy for a single-user link that reduces reliance on full channel knowledge by exploiting stable geometric features of the propagation field. In particular, the beam is steered toward a principal propagation path that remains relatively stable over time. Experimental results using the SPACE and MACE data sets demonstrate reliable communication with excellent data-detection mean-squared error and zero bit errors.

Top

 Technical Session: “Physical Layer: Modulation and Signal Processing”

Authors: T. Corner, J. Neasham

Abstract: Multi-element detection methods are investigated that exploit the spatial channel diversity of a Vertical Line Array (VLA) to maximize the frame synchronization reliability of an underwater acoustic receiver. These methods improve data throughput in negative Signal-to-Noise Ratio (SNR) underwater acoustic channels. The detection methods are evaluated on sea trial data captured in the North Sea using a 5-element VLA with 1m element spacing. Signals were transmitted in two bands: 8-12kHz and 20-28kHz. For both bands the aperture of the array is greater than many wavelengths, so planar wave beamforming methods are not applicable. Non-coherent combination methods have been investigated to create a multi-element detector which increases data throughput over a single-element detector by up to 70% in low SNR, spatially diverse channels, with no additional packet overheads. Results show that a combination of detection methods are required to optimally exploit the variation of channels present in underwater acoustic communication.

Authors: G. Chua, M. Chitre

Abstract: Underwater acoustic receivers often separate channel estimation, equalization, and forward-error correction (FEC) into successive stages, which limits how pilot observations and code constraints can support one another during packet recovery. This paper proposes Differentiable Forward Error Correction (DFEC), a receiver that refines an effective channel impulse response and a soft transmitted codeword within a single differentiable objective. The objective combines a signal least-squares term, a soft low-density parity-check (LDPC) parity constraint, and a pilot penalty applied to known codeword coordinates. These pilot constraints can participate in the objective according to the LDPC parity matrix. This differs from an iterative Turbo Equalizer (TurboEQ): TurboEQ exchanges soft information between equalizer and decoder, but the pilot-supported channel or equalizer estimate is formed upstream rather than refined inside the same pilot–code–channel objective. We evaluate DFEC on BPSK packets from an underwater field experiment and compare it with TurboEQ and a decision-feedback equalizer (DFE) followed by FEC, under matched pilot budgets. The results show that the proposed joint approach improves packet success rate (PSR) relative to TurboEQ and DFE+FEC.

Authors: P. Donnelly, K. Enhos, E. Demirors, T. Melodia

Abstract: The underwater acoustic (UWA) channel is among the most adverse known to wireless communication research. Due to its doubly-selective and highly-reflective nature, the UWA channel exhibits dense multipath effects, introducing significant inter-symbol interference (ISI) in all but the most sparsely populated environments, necessitating robust channel equalization. In this paper, we implement a variant of the iterative multichannel equalization and decoding (IMED) algorithm, a multi-receiver implementation of decision-feedback turbo equalization (DFTE), using the Hydronet Nexus software-defined modem and edge platform, and provide what we believe to be the first examination of its effectiveness over a non-line-of-sight high-multipath UWA channel. We show an average improvement in BER of one order of magnitude per additional receiver at reasonable SNR levels over the challenging 160-meter UWA link, with an average equalized PER as low as 17.8%, enabling burgeoning maritime applications such as environmental monitoring, infrastructure inspection, and unmanned underwater vehicle operations through significantly adverse channels.

Authors: P. van Walree, M. Chitre, F. Socheleau

Abstract: This paper illustrates the versatility of SODAC, an open access dataset with channel probes and generic communication waveforms recorded in a reverberant environment. It quantifies the reverberation by means of a cumulative power function and explores several ideas to deal with its detrimental impact on communication systems. Forward error coding and repetition coding are applied to improve throughput at the expense of spectral efficiency. Time diversity processing and successive interference cancellation yield promising results that preserve spectral efficiency. A reuse example of communication in a non-cooperative context is provided through blind parameter estimation of waveform parameters.

Authors: M. Deguchi, Y. Kida

Abstract: In underwater acoustic communications for autonomous underwater vehicles (AUVs), reliable data transmission is essential under time-varying channel conditions. Decision feedback equalizers (DFEs) based on the recursive least squares (RLS) algorithm are widely used to mitigate severe multipath effects. However, the demodulation performance by the conventional RLS degrades significantly in the presence of strong inter-channel correlation, often arising in compact receiver arrays. A covariance block-diagonalized RLS (CBD-RLS) has been proposed, which improves demodulation performance while reducing computational complexity by neglecting inter-channel correlation. Despite its empirical success, the underlying mechanism remains unclear. This paper analyzes the instability of conventional RLS under strong inter-channel correlation and clarifies how CBD-RLS achieves robust estimation. We show that strong inter-channel correlation leads to ill-conditioned covariance matrices, resulting in variance amplification and numerical instability in RLS, whereas CBD-RLS suppresses this effect at the cost of introducing a correlation-induced bias. Experimental results using underwater acoustic data demonstrate that inter-channel correlation induces rank deficiency in the covariance matrix for RLS, while CBD-RLS maintains a high effective rank through covariance block diagonalization, leading to stable estimation.

Top

Technical Session: “Quantum and Optical Solutions for the Maritime Domain”

Authors: S. Gladysz, P. Paglierani

Abstract: We present a comprehensive theoretical framework that can be used to predict the performance of various quantum key distribution protocols for secure data transmission in underwater optical communications. We assume that both quantum and conventional transmission are carried by visible-light laser beams. Using this framework, it is possible to conduct trade-off studies on factors such as the diameters of the transmitting and receiving apertures, wavelengths, the receiver’s field of view, and detection modes (e.g. single-mode fiber versus free-space detector).

Authors: J.F. Bousquet, S. Wang, M. Saber

Abstract: This work describes a magnetic sensor to cross the seawater interface for a magneto-inductive wireless link to an underwater node. The receiver is enhanced with a high-sensitivity Superconducting Quantum Interference Device (SQUID). Constrained on the transmitter characteristics, the magnetic field intensity is calculated at the receiver as a function of distance and frequency. This allows to provide the sensor electronic front-end requirements. Specifically, the nonlinear transfer function is obtained between the receiver input field intensity and the SQUID output voltage. The receiver includes a Flux Locked Loop (FLL) to maintain the circuit stable in presence of noise and linear for distances ranging between 5 meter to 30 meters.

Authors: I. B. Saksvik, V. Hassani

Abstract: Optical wireless underwater communication (OWUC) currently lacks an open communication standard. As a result, optical modems often function as wireless adapters that forward the communication stack from a wired cable (e.g., Ethernet/IP packets), resulting in unnecessary protocol overhead. To address this, we present PROTEUS, a lightweight physical- and data-link-layer protocol for point-to-point (P2P) optical links. PROTEUS consists of a small 15-byte total frame overhead and supports user data of up to 1 KB. We validate the protocol through sea trials of LED-based optical modems in the Oslofjord, Norway, where a 1 MB file was transmitted between a static and a moving platform at 2 Mbps. PROTEUS does not claim any novel protocol mechanisms. Rather, it reuses framing conventions common to existing digital communication protocols and represents, to the best of the authors’ knowledge, the first effort toward an OWUC protocol in the current literature.

Top

Technical Session: “Security in Underwater Communications”

Authors: D. Eccher, F. Ferreira, P. Casari

Abstract: The interest for underwater acoustic networks in critical applications is increasing the efforts towards realizing reliable security functions. In this paper, we consider a node joining a network, and propose a physical layer security approach to detect sybil attacks attempting to affect the joining process. Our solution involves extracting the channel impulse response (CIR) from the signal received from each node that answers the join request, including legitimate nodes and impersonating attackers. Through a mismatch metric that relies on the spatial de-correlation of underwater acoustic channels, we propose an approach to identify the nodes that yield similar CIRs, and are thus unlikely to be legitimate. Simulation results show that our method successfully tells legitimate and fake nodes apart, and yields a high probability that the final set of chosen trusted nodes contains a majority of legitimate network devices.

Authors: G. Gubnitsky, R. Diamant

Abstract: Traditional underwater acoustic release systems rely on predefined commands or static credentials that require persistent storage, making them vulnerable to extraction through device compromise or physical capture and enabling unauthorized activation. We propose a secure authentication framework that avoids storing secrets by coupling user identity with the instantaneous acoustic channel. The method reconstructs a biometric key at the topside modem (Alice) using a fuzzy extractor, while the subsea node (Bob) retains only a protected hash of the enrolled biometric template. Following a probe exchange, both parties independently extract two robust channel features: the propagation delay and the arrival delay between the direct and surface reflected paths, which are quantized to derive a session-specific channel key. The final authentication key is obtained by hashing the reconstructed biometric key with the channel-derived key, ensuring that authentication is both user- and location-specific and cannot be reused across deployments. The system is evaluated using 10 million Monte Carlo simulations of underwater acoustic channels and a fingerprint dataset comprising 51 identities. Performance is assessed via Hamming distance distributions and receiver operating characteristic analysis. Results demonstrate strong resilience to impersonation and replay attacks, achieving reliable discrimination at signal-to-noise ratios as low as −23 dB using a simple matched-filter receiver.

Authors: Z. Li, D. Pompili

Abstract: This paper proposes Key-Agile Time-Shift Signaling (KATS), a waveform-embedded keyed secrecy design for cyclic time-shift signaling in underwater acoustic communication. KATS makes the waveform root and cyclic alignment secret-driven and frame-varying, preserving low-complexity noncoherent modulation while forcing unauthorized correlator-based receivers onto a mismatched template. The proposed design separates transmission into a minimal public coordination header and a secret time-shift payload layer recoverable only by authorized receivers with the correct secret parameters. We analyze the mismatched-template behavior of unauthorized receivers and show that incorrect waveform hypotheses destroy the correspondence between dominant correlation peaks and payload symbols, leading to “confidently wrong” decoding behavior. Performance evaluation using realistic underwater acoustic channels from the ACommSet dataset shows that KATS achieves substantially lower attacker information leakage than chaotic DSSS/CDMA baselines while preserving reliable noncoherent communication performance.

Authors: B. Krivokapic, S. Tomovic, I. Radusinovic, J. Neasham

Abstract: Bio-mimetic covert underwater acoustic communication hides transmissions by making them statistically indistinguishable from marine biological sounds. Existing contour-modification schemes embed payloads via hand-designed, whistle-agnostic perturbation rules that introduce periodic frequency artefacts detectable by statistical tests. We propose a neural contour-steganography system in which a compact convolutional encoder embeds a binary payload as a learned, bounded frequency perturbation on a real dolphin-whistle contour, conditioned jointly on the payload and the host contour. An adversarial discriminator and smoothness regularisation keep perturbed contours within the natural whistle manifold. A two-stage training procedure jointly trains encoder and decoder against synthetic contour corruptions to establish covertness, then freezes the encoder and adapts only the decoder to the full simulated acoustic pipeline, preserving covertness while closing the gap to deployment conditions. Evaluated on real dolphin recordings, the proposed system matches two rule-based contour-modulation baselines in decoding accuracy across 5- and 8-bit payloads, while producing perturbed contours that are statistically closer to natural dolphin whistles and significantly harder to detect by a learned classifier.

Underwater Communications and Networking Conference
Powered by  GDPR Cookie Compliance
Privacy Overview

More information about our Privacy Policy