This paper investigates range-depth localization of an orthogonal frequency-division multiplexing (OFDM) source using a single hydrophone, which serves as a portable scheme suitable for small underwater platforms. We consider a typical shallow-water propagation model, where the rich position information embedded in multipath delays and amplitudes offers the potential to achieve the above purpose. Specifically, we exploit the frequency-domain input-output relationship of pilot subcarriers to construct a measurement model and incorporate joint processing of multiple OFDM symbols to enhance localization performance. To characterize random fluctuations in path gains caused by scattering from the rough surface and bottom, we introduce the Rician fading model. Under this assumption, each complex path amplitude comprises a deterministic component associated with the source position and a random perturbation. To fully leverage the parametric model, we adopt the stochastic maximum likelihood principle. The resulting high-dimensional optimization problem is iteratively solved by the developed expectation-maximization-type algorithm. In addition, a stable initialization scheme is designed to strive for convergence to the global optimum. To quantitatively evaluate the sensing capability of the integrated system, we derive the Cramér-Rao bound on the mean square error of the localization result and elucidate the performance gain from the joint processing. For completeness, we also present a tailored communication processing chain for symbol detection. Simulations and laboratory experiments demonstrate the superiority of the proposed algorithm over existing matched field processing-based positioning techniques.

End-to-end (E2E) delay analysis is a prerequisite for protocol and algorithm design in water-air cross-media heterogeneous networks (WACHN). However, the heterogeneity of protocols and channels poses a significant challenge, as existing analytical methods cannot accommodate such diverse models. To address this gap, this work proposes a stochastic network calculus (SNC) based analytical framework for E2E delay in WACHN. Specifically, radio and underwater acoustic channel performance is characterized by modeling the relationship between channel quality and bit error rate using the three-ray and Thorp models. Furthermore, a Markov state transition service model is introduced for the underwater acoustic fountain code-based automatic repeat request mechanism, accounting for packet fragmentation, redundancy, and non-traffic conservation characteristics for the first time. Building on these models, this work quantifies the coupling effects of media switching, packet size variations, and slot-scale heterogeneity on E2E delay within the proposed SNC framework. Subsequently, a closed-form expression for the delay’s complementary cumulative distribution function (CCDF) is derived using martingale theory. To validate the theoretical derivation, Monte Carlo simulations demonstrate that analytical results accurately capture key statistical features such as distribution shape and tail behavior, validating the derived CCDF expression’s ability to provide a strict and effective delay upper bound. Consequently, the proposed framework establishes the first theoretical foundation for performance analysis and optimization of WACHN, comprehensively addressing media disparities.

Long-term ice dynamics of the Prince Harald system in Lützow-Holm Bay (LHB), East Antarctica, particularly before 1990, remain poorly investigated, hindering assessments of its historical contribution to regional mass balance. Here, we reconstructed high-resolution ice velocity fields (1973-1989) for its main components – Prince Harald 1, Prince Harald 2, and Prince Harald 3 – leveraging historical Landsat imagery with a robust photogrammetric method, a systematic framework that includes geometric correction, orthorectification, and hierarchical feature matching. Our integrated analysis, combining these velocities with assessments of basal melt, surface elevation change, ice front dynamics, and Passive Shelf Ice (PSI) behavior, reveals a stable state of the system. We find that this stability was characterized by consistent flow velocities, persistent surface thickening, minimal basal melt, and calving events predominantly within the PSI zone. These results indicate an overall trend of stability and net mass accumulation for the Prince Harald system during this period.

The multi-channel outputs of an acoustic vector sensor (AVS) provide diversity gain for communications, and developing an effective combining scheme becomes a critical issue. However, in underwater multipath channels, a single AVS struggles to estimate the spatial signatures of multipath signals due to its limited sensing capability, which compromises the design of optimal combining weights. To overcome this issue, we propose a location-aided maximal ratio combining (MRC) technique. Armed with a predictable end-to-end propagation model, we first develop a maximum-likelihood sensing framework with the help of the pilot subcarriers embedded in the OFDM signal. The required channel state information is inferred from the estimated propagation geometry. Then, the combining weight vector is determined according to the MRC principle. Simulations demonstrate that this integrated scheme enhances communication performance through comprehensive environmental sensing.

This article investigates a direction-of-arrival (DOA) estimation method for underwater acoustic arrays in non-Gaussian impulsive noise environments. Traditional DOA estimation methods for underwater acoustic arrays typically presume that underwater environmental noise follows a Gaussian distribution. This assumption can lead to a significant degradation in estimation accuracy, or even failure, in underwater environments where non-Gaussian impulsive noise is predominant, thereby limiting the detection capabilities of underwater acoustic arrays. To address this issue, this study employs a method based on mixture correntropy, which maximizes the mixture correntropy of the residual fitting error matrix for subspace decomposition of the received data matrix, effectively filtering out impulsive noise. Considering the signal processing performance of correntropy and mixture correntropy depends on the selection of the kernel width, this article introduces a novel adaptive method for updating the kernel width. This method updates the kernel width in each iteration based on the residual fitting error value, setting the square of the kernel width to the sum of the squares of a preset kernel width and the residual fitting error modulus. This approach retains the simplicity and robustness of the maximum mixture correntropy criterion (MMCC) algorithm while enhancing the convergence rate and achieving a lower steady-state excess mean square error. Furthermore, this study applies the classical multiple signal classification (MUSIC) algorithm for DOA estimation. Finally, simulations and sea trials have substantiated the correctness and effectiveness of the method proposed in this article.

The laser altimetry data acquired from the Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) are instrumental in monitoring the Antarctic ice sheet (AIS) and glaciers. Its land ice elevation product (ATL06) has demonstrated an impressive elevation accuracy of 2–4 cm over flat inland ice areas. However, under the complex and extreme conditions of ice sheet marginal regions, a more comprehensive understanding of ICESat-2’s geolocation capability is crucial. Such insights are essential for enhancing the satellite’s ability to accurately quantify dynamic changes in Antarctic glaciers and marginal zones, thereby supporting precise regional mass balance estimations. During the 36th Chinese National Antarctic Research Expedition (CHINARE) in 2019–2020, high-precision autonomous aerial vehicle (AAV) surveys supported by the Global Navigation Satellite Systems (GNSS) were conducted near ICESat-2 overpass times to systematically evaluate and improve ICESat-2’s geolocation capability in Antarctic marginal regions. Our analysis revealed that the ATL06 data’s native 20-m resolution often lacks the granularity to resolve fine topographic features in these areas. ICESat-2 photon data (ATL03), however, hold the potential for more accurate elevation estimations, particularly in the photon-counting mode on complex ice surfaces. To address this, we propose a framework for evaluating and improving ICESat-2 photon data geolocation capability based on high-precision AAV digital surface model (DSM) registration. This framework uses a “photon cloud” profile fitting method on ATL03 photon data, enhancing the elevation profile by selectively filtering noise and applying adaptive polynomial fitting. Geolocation performance of the resulting high-resolution (0.7 m) ICESat-2 elevation profiles was evaluated by analyzing discrepancies between ICESat-2 profiles and AAV DSM data, considering geolocation error, terrain slope, and curvature, along the ice sheet margin near Zhongshan Station, East Antarctica. Geolocation offsets for ICESat-2 range from 2.20 to 5.40 m. Our results demonstrate that the proposed framework improves the ICESat-2 photon data’s ability to monitor dynamic ice features and surface elevation changes in Antarctic marginal regions.

The development in machine learning (ML) technology has brought new horizons for the prediction of subglacial lakes (SLs) using radio-echo sounding (RES) data, offering fresh perspectives toward the automated identification of SLs. Nonetheless, the inherent data imbalance across various classes within the dataset presents significant analytical challenges. To address this limitation, the artificial bee colony (ABC) optimization algorithm is introduced to automatically predict SLs in Gamburtsev Province in East Antarctica, using an optimized stacking ensemble learning approach. The proposed method predicts SLs by using five representative features selected through importance and correlation analyses of eight features derived from RES data. The experimental outcomes demonstrate the superiority of this method in overcoming the significant imbalance of RES data, successfully identifying known lakes in the validation dataset. Furthermore, this study summarizes an inventory of SLs across the Gamburtsev subglacial mountains in East Antarctica, and a total of 55 new candidate SLs with lengths ranging from 108 to 38130 m have been predicted using our novel method. The source code is publicly available athttps://github.com/vivian-ma97/ABC-Stacking-for-Subglacial-Lakes

For the purpose of computational efficiency, we propose two subspace-based methods, but without eigendecomposition, to address the two typical problems in nested array processing, i.e., direction-of-arrival (DOA) estimation and noise elimination. In detail, to estimate DOA parameters, we judiciously arrange the segments extracted from the co-array model and then introduce a novel co-array-based orthogonal propagator method (COPM). Next, we develop a projection-based noise cancellation approach in the co-array domain, improving the relatively poor performance of COPM at low signal-to-noise ratios. Simulations evaluate the proposed algorithms under both overdetermined and underdetermined conditions.

The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2), equipped with the Advanced Topographic Laser Altimetry System (ATLAS), enhances ice surface elevation change estimations. ATLAS significantly improves elevation accuracy from over 10 cm, as seen in ICESat’s full-waveform technique, to 2-4 cm. ICESat-2 data’s elevation quality is currently validated at the single-pulse level using the corner cube retro-reflector (CCR) method, a terrain-independent technique insensitive to ice surface conditions. This study introduces a cost-effective CCR system for validating ICESat-2 elevation data from Antarctica’s firn surface, complementing the ICESat-2 Science Team’s approach. Through three validation campaigns in Antarctica and Shanghai, China, we enhanced our CCR system in design of the prism hardware, alignment of the CCR array, and determination of valid photon data. We implemented a self-adaptive window determination strategy for valid CCR-returned signal photons to minimize the impact of inadequately returned photons from the Fraunhofer diffraction pattern’s lobes and two closely elevated CCRs. The accurate elevations of single pulses from each CCR were determined and analyzed using Global Navigation Satellite Systems (GNSS) in-situ observations during the ICESat-2 overpasses in 2020–2022. Our refined CCR validation system shows that ICESat-2’s pulse elevation has a bias of less than 3.0 cm and a precision of ±1.7 cm in the latest experiment on the Nansen ice shelf, East Antarctica.

Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) carries a photon-counting laser altimeter with an unprecedented elevation accuracy of 2–4 cm. Since its data availability in 2018, there has been a challenge for the establishment of a new data processing model that can take advantage of this satellite for accurately estimating volumetric changes in Antarctica and associated contribution to global sea level rise (GSLR). We introduce an innovative multitemporal elevation change estimation model (MECEM) that separates precipitation effects from topographic influences to eliminate their correlations and estimates the elevation change rates effectively through a spatiotemporal iterative procedure. The MECEM results are validated by using GNSS in situ observations, snow stakes measurements, and airborne altimetric survey data. The results are also compared with those from ICESat and ESA multimission radar altimetric dataset. It is demonstrated that the model is capable of estimating small thickening of$1.8~\pm ~0.1$cm yr1 in the Vostok subglacial lake region. Using ICESat-2 ATL06 data from 2019 to 2023, the model is proven to be effective in the estimation of elevation change rates in Antarctic basins of different characteristics. Our results show that an increase of$0.103~\pm ~0.001$m yr1 in thickening is found from 2017–2021 to 2019–2023 in Dronning Maud Land. Furthermore, an accelerated thinning by$- 0.12~\pm ~0.035$m yr1 is witnessed from 2003–2019 to 2019–2023 in the fast-flowing Pine Island Glacial. With more ICESat-2 data acquired, the developed MECEM model can be applied for estimating the contribution of the entire Antarctic ice sheet (AIS) to GSLR.

With the proliferation of marine activities, underwater Internet of Things (UIoT), which integrates various techniques for supporting smart ocean, has attracted more research interest. The trend is that one desires to use computationally efficient and widely applicable algorithms for source localization. For fulfilling the above requirements, this paper proposes a novel unitary quaternion (UQ) model, which is applied to widespread centro-symmetric arrays. The estimation and decomposition of the corresponding covariance matrix can be executed in the real number field, thus benefiting from low complexity. Moreover, we analyze the physical implication of the proposed model and associate it with the emerging quaternion-based attitude estimation and control, which reveals the potential advantages of the UQ model in UIoT. In the simulations, we test the algorithm performance in several realistic underwater scenarios, demonstrating the flexibility and applicability of the UQ model.

Nested acoustic vector sensor (AVS) arrays have attracted growing interest, and their performance can be further improved by assembling AVSs with spatially separated (SS) configurations. Following this trend, we propose an augmented tensor-MUSIC (AT-MUSIC) method for SS-AVSs while exploiting the inherent multidimensional structure of the array data. The implications of “augmented” are twofold: (i) conventional tensor models based on co-centered AVSs are generalized to the case of SS-AVSs, and (ii) available information is fully utilized by judiciously arranging and aggregating the received data. Finally, we show the enhanced performance of AT-MUSIC in simulations.

Being able to efficiently and accurately select the top-$k$ elements with differential privacy is an integral component of various private data analysis tasks. In this paper, we present the oneshot Laplace mechanism, which generalizes the well-known Report Noisy Max \cite{dwork2014algorithmic} mechanism to reporting noisy top-$k$ elements. We show that the oneshot Laplace mechanism with a noise level of $\widetilde{O}(\sqrt{k}/\eps)$ is approximately differentially private. Compared to the previous peeling approach of running Report Noisy Max $k$ times, the oneshot Laplace mechanism only adds noises and computes the top $k$ elements once, hence much more efficient for large $k$. In addition, our proof of privacy relies on a novel coupling technique that bypasses the composition theorems so without the linear dependence on $k$ which is inherent to various composition theorems. Finally, we present a novel application of efficient top-$k$ selection in the classical problem of ranking from pairwise comparisons.

In recent years, human trajectory prediction (HTP) has garnered attention in computer vision literature. Although this task has much in common with the longstanding task of crowd simulation, there is little from crowd simulation that has been borrowed, especially in terms of evaluation protocols. The key difference between the two tasks is that HTP is concerned with forecasting multiple steps at a time and capturing the multimodality of real human trajectories. A majority of HTP models are trained on the same few datasets, which feature small, transient interactions between real people and little to no interaction between people and the environment. Unsurprisingly, when tested on crowd egress scenarios, these models produce erroneous trajectories that accelerate too quickly and collide too frequently, but the metrics used in HTP literature cannot convey these particular issues. To address these challenges, we propose (1) the A2X dataset, which has simulated crowd egress and complex navigation scenarios that compensate for the lack of agent-to-environment interaction in existing real datasets, and (2) evaluation metrics that convey model performance with more reliability and nuance. A subset of these metrics are novel multiverse metrics, which are better-suited for multimodal models than existing metrics. The dataset is available at: https://mubbasir.github.io/HTP-benchmark/.

In this letter, we present a novel computationally efficient DOA estimation algorithm based on quaternion theory for two-dimensional (2-D) direction-of-arrival (DOA) estimation. An orthogonal propagator method based on the cross-correlation of the quaternion models (OPM-CQM) is developed to alleviate the computation burden. To eliminate the effect of additive noise, we construct two quaternion-based signal models judiciously. Then, we obtain the statistics of the observed signals by performing the cross-correlation between the quaternion models. Meanwhile, the additive noise is eliminated without introducing other denoising methods. Moreover, the compact modeling approach based on quaternions provides a significant advantage to OPM-CQM in terms of computational effort. Simulations demonstrate that the proposed algorithm offers performance superiority in angular resolution compared with the non-quaternion schemes.

This letter proposes a correlation detection algorithm based on multi-channelization. The existing method has poor performance for detecting a time-varying signal with a low signal-to-noise ratio(SNR). To overcome this problem, we use the method of correlation summation after multi-channelization. According to the difference between the correlation between signals and noise in the received data, the single-channel data is converted into multi-channel data by a multi-channelization method without changing the signals. Then the noise is filtered by correlation calculation of multi-channel data, which can improve the detection probability of signals under low SNR. Based on the theory of multi-channelization, we propose two different methods to detect signals. To better reflect the performance of the algorithm, we compared several classical signal detection methods, such as energy detection (ED), correlation detection (CD), and cyclic spectrum detection (CSD). Simulation analysis and experimental results show that the multi-channelization correlation detection algorithm has better detection performance when the SNR is low.

Historical optical photographs acquired by CORONA, ARGON and LANYARD missions, known as Declassified Intelligence Satellite Photography (DISP), date back to the satellite remote sensing in Antarctica in early 1960s. They can provide a broader perspective for studying the early Antarctica by recovering geophysical parameters and products. However, the quality of the early optical satellite images is relatively poor due to the imaging technique used, long-term storage of the film and analogue-to-digital scanning processes. Improvements should be applied to processing methods to make these valuable images suitable for analyzing changes in the Antarctica. In this paper, we introduced our methods of geometric position, hierarchical image matching and parallax decomposing, then demonstrated that these images can be successfully applied to applications, such as DEM reconstruction and ice flow mapping.

Crowd simulation, the study of the movement of multiple agents in complex environments, presents a unique application domain for machine learning. One challenge in crowd simulation is to imitate the movement of expert agents in highly dense crowds. An imitation model could substitute an expert agent if the model behaves as good as the expert. This will bring many exciting applications. However, we believe no prior studies have considered the critical question of how training data and training methods affect imitators when these models are applied to novel scenarios. In this work, a general imitation model is represented by applying either the Behavior Cloning (BC) training method or a more sophisticated Generative Adversarial Imitation Learning (GAIL) method, on three typical types of data domains: standard benchmarks for evaluating crowd models, random sampling of state-action pairs, and egocentric scenarios that capture local interactions. Simulated results suggest that (i) simpler training methods are overall better than more complex training methods, (ii) training samples with diverse agent-agent and agent-obstacle interactions are beneficial for reducing collisions when the trained models are applied to new scenarios. We additionally evaluated our models in their ability to imitate real world crowd trajectories observed from surveillance videos. Our findings indicate that models trained on representative scenarios generalize to new, unseen situations observed in real human crowds.

Resource constraints frequently complicate multi-agent planning problems. Existing algorithms for resource-constrained, multi-agent planning problems rely on the assumption that the constraints are deterministic. However, frequently resource constraints are themselves subject to uncertainty from external influences. Uncertainty about constraints is especially challenging when agents must execute in an environment where communication is unreliable, making on-line coordination difficult. In those cases, it is a significant challenge to find coordinated allocations at plan time depending on availability at run time. To address these limitations, we propose to extend algorithms for constrained multi-agent planning problems to handle stochastic resource constraints. We show how to factorize resource limit uncertainty and use this to develop novel algorithms to plan policies for stochastic constraints. We evaluate the algorithms on a search-and-rescue problem and on a power-constrained planning domain where the resource constraints are decided by nature. We show that plans taking into account all potential realizations of the constraint obtain significantly better utility than planning for the expectation, while causing fewer constraint violations.

Ice flow velocity is used to estimate ice mass changes in glaciers and is a significant indicator of the stability of the Antarctica ice sheet in global change studies. The existing regional Antarctica ice flow speed maps are usually derived from radar or optical satellite observations of modern satellites since the 1970s. This paper presents a new analytical photogrammetric method for estimating Antarctica ice flow velocity fields by using film-based stereo ARGON photographs collected in the 1960s. The key of the proposed innovative method is a parallax decomposition that separates the effect of the terrain relief from the ice flow motion. An innovative implementation strategy is developed by using a framework that involves key techniques of hierarchical stereo image matching, ice flow direction determination, parallax decomposition, and ice flow speed estimation. This method is applied in the Rayner glacier in eastern Antarctica by using two sets of ARGON images with a two-month interval in 1963. The produced digital terrain model and speed map achieved a ground position accuracy of 61 m and a speed accuracy of 70 m a-1. A comparison with recent products from 2000 to 2010 shows no significant topographic changes in the study area. Furthermore, the speed around the grounding line remained at the same level, while the speed in the ice shelf front decreased by 73 m a-1. The ice shelf front advanced by approximately 7 km over more than 40 years. Overall, the observation results indicate favorable conditions for the stability of the Rayner glacier-ice shelf system.