Reconstructing time-varying graph signals (or graph time-series imputation) is a critical problem in machine learning and signal processing with broad applications, ranging from missing data imputation in sensor networks to time-series forecasting. Accurately capturing the spatio-temporal information inherent in these signals is crucial for effectively addressing these tasks. However, existing approaches relying on smoothness assumptions of temporal differences and simple convex optimization techniques that have inherent limitations. To address these challenges, we propose a novel approach that incorporates a learning module to enhance the accuracy of the downstream task. To this end, we introduce the Gegenbauer-based graph convolutional (GegenConv) operator, which is a generalization of the conventional Chebyshev graph convolution by leveraging the theory of Gegenbauer polynomials. By deviating from traditional convex problems, we expand the complexity of the model and offer a more accurate solution for recovering time-varying graph signals. Building upon GegenConv, we design the Gegenbauer-based time graph neural network (GegenGNN) architecture, which adopts an encoder–decoder structure. Likewise, our approach also uses a dedicated loss function that incorporates a mean squared error (MSE) component alongside Sobolev smoothness regularization. This combination enables GegenGNN to capture both the fidelity to ground truth and the underlying smoothness properties of the signals, enhancing the reconstruction performance. We conduct extensive experiments on real datasets to evaluate the effectiveness of our proposed approach. The experimental results demonstrate that GegenGNN outperforms state-of-the-art methods, showcasing its superior capability in recovering time-varying graph signals.

Merchant ship-radiated noise, recorded on a single receiver in the 360–1100 Hz frequency band over 20 min, is employed for seabed classification using an ensemble of deep learning (DL) algorithms. Five different convolutional neural network architectures and one residual neural network are trained on synthetic data generated using 34 seabed types, which span from soft-muddy to hard-sandy environments. The accuracy of all of the networks using fivefold cross-validation was above 97%. Furthermore, the impact of the sound speed and water depth mismatch on the predictions is evaluated using five simulated test cases, where the deeper and more complex architectures proved to be more robust against this variability. In addition, to assess the generalizability performance of the ensemble DL, the networks were tested on data measured on three vertical line arrays in the Seabed Characterization Experiment in 2017, where 94% of the predictions indicated that mud over sand environments inferred in previous geoacoustic inversions for the same area were the most likely sediments. This work presents evidence that the ensemble of DL algorithms has learned how the signature of the sediments is encoded in the ship-radiated noise, providing a unified classification result when tested on data collected at-sea.

The presence of internal waves (IWs) in the ocean alters the isotropic properties of sound speed profiles (SSPs) in the water column. Changes in the SSPs affect underwater acoustics since most of the energy is dissipated into the seabed due to the downward refraction of sound waves. In consequence, variations in the SSP must be considered when modeling acoustic propagation in the ocean. Empirical orthogonal functions (EOFs) are regularly employed to model and represent SSPs using a linear combination of basis functions that capture the sound speed variability. A different approach is to use dictionary learning to obtain a learned dictionary (LD) that generates a nonorthogonal set of basis functions (atoms) that generate a better sparse representation. In this paper, the performance of EOFs and LDs are evaluated for sparse representation of SSPs affected by the passing of IWs. In addition, an LD-based supervised framework is presented for SSP classification and is compared with classical learning models. The algorithms presented in this work are trained and tested on data collected from the Shallow Water Experiment 2006. Results show that LDs yield lower reconstruction error than EOFs when using the same number of bases. In addition, overcomplete LDs are demonstrated to be a robust method to classify SSPs during low, medium, and high IW activity, reporting accuracy that is comparable to and sometimes higher than that of standard supervised classification methods.

Graph Neural Networks (GNNs) have shown great promise in modeling relationships between nodes in a graph, but capturing higher-order relationships remains a challenge for large-scale networks. Previous studies have primarily attempted to utilize the information from higher-order neighbors in the graph, involving the incorporation of powers of the shift operator, such as the graph Laplacian or adjacency matrix. This approach comes with a trade-off in terms of increased computational and memory demands. Relying on graph spectral theory, we make a fundamental observation: the regular and the Hadamard power of the Laplacian matrix behave similarly in the spectrum . This observation has significant implications for capturing higher-order information in GNNs for various tasks such as node classification and semi-supervised learning. Consequently, we propose a novel graph convolutional operator based on the sparse Sobolev norm of graph signals. Our approach, known as Sparse Sobolev GNN (S2-GNN), employs Hadamard products between matrices to maintain the sparsity level in graph representations. S2-GNN utilizes a cascade of filters with increasing Hadamard powers to generate a diverse set of functions. We theoretically analyze the stability of S2-GNN to show the robustness of the model against possible graph perturbations. We also conduct a comprehensive evaluation of S2-GNN across various graph mining, semi-supervised node classification, and computer vision tasks. In particular use cases, our algorithm demonstrates competitive performance compared to state-of-the-art GNNs in terms of performance and running time.

El cambio climático global está afectando a la Tierra de diversas maneras, como el desprendimiento de glaciares, la subida del nivel del mar y olas de calor más intensas, que a su vez contribuyen a los incendios en zonas ricas en flora y fauna, provocando erosión y cambios en las áreas de distribución de plantas y animales. A pesar de los esfuerzos internacionales por mitigar el cambio climático, el problema sigue sin controlarse en muchas zonas. En este estudio se propone un sistema de vigilancia y alerta temprana de incendios forestales. Se basa en una red de sensores inalámbricos (WSN) que mide el entorno para detectar y prevenir incendios en zonas rurales. La WSN utiliza tecnologías de acceso abierto y el estándar Zigbee para la comunicación por radiofrecuencia y envía los datos recogidos a un servidor web mediante GPRS y protocolo TCP/IP. El rendimiento de la WSN se evalúa utilizando dos métricas: la tasa de transmisión de datos y el retardo temporal. La WSN logró una tasa media de éxito en la transmisión de datos superior al 86% con un retardo medio inferior a 500 milisegundos en todas las pruebas, lo que demuestra el potencial de las WSN como sistemas de detección de incendios forestales casi en tiempo real.

Conventional techniques for underwater source localization have traditionally relied on optimization methods, matched-field processing, beamforming, and, more recently, deep learning. However, these methods often fall short to fully exploit the data correlation crucial for accurate source localization. This correlation can be effectively captured using graphs, which consider the spatial relationship among data points through edges. This work introduces a novel graph learning module for source localization using spectrograms from ships-of-opportunity, which represent mid-frequency acoustic broadband signals from ship-radiated noise ranging from 360 to 1100 Hz, collected during the 2017 Seabed Characterization Experiment (SBCEX 2017). The proposed approach follows a two-step process: first, a pre-trained convolutional neural network (CNN) module is used for feature extraction via self-supervised learning, and then a graph neural network model is trained using semi-supervised learning for source localization. The graph is constructed using a k-nearest neighbor algorithm, incorporating features extracted by the CNN from the spectrograms. By employing this two-stage training strategy, our framework addresses the challenge of limited labeled data availability while achieving performance comparable to conventional supervised learning models. The effectiveness of our approach is demonstrated through model evaluation on both synthetic and measured data, showcasing the architecture's ability to generalize well to unseen scenarios.