The integrated space-air-ground-sea communication network has attracted increasing attention in recent years. Among its challenges, across air-sea medium communication remains a critical bottleneck due to its narrow bandwidth and severe attenuation. In this study, we investigate across-medium detection of acoustic-induced surface waves excited by encoded and modulated underwater acoustic signals, leveraging the characteristics of both underwater acoustic and millimeter-wave channels. Binary information carried by acoustic signals is used to generate surface waves at the water-air interface, which are directly detected in the air via millimeter-wave radar. After filtering under conditions of strong background noise and multi-source acoustic interference, the original acoustic information is accurately extracted. Experimental validation and simulation results demonstrate that the amplitude of 2FSK-modulated acoustic-induced surface waves detected across the air-sea interface ranges from 0.8522 μm to 1.0758 μm, with a bit error rate (BER) of 0. Out-of-band signal amplitudes are suppressed to 0.239 μm. Under identical symbol durations, the multi-channel scheme proportionally increases the across-medium communication capacity and data rate. In complex maritime noise environments, comparison of single-channel, dual-channel, and four-channel 2FSK schemes shows that even when the transmission rate is increased to 200 bps using four channels, millimeter-wave radar can still reliably detect the acoustic-induced surface waves with a BER below 3.2%. This study enables direct airborne detection and high-accuracy decoding of binary information encoded in underwater acoustic signals under complex sea conditions, demonstrating significant potential for practical implementation in high-speed uplink communication systems across the air-sea interface.

In many scientific domains involving one-dimensional data processing, such as spectral analysis, classical models like Principal Component Analysis (PCA) are favoured for their interpretability but fail to capture nonlinear structures. Conversely, the ‘black box’ nature of deep learning impedes its adoption where trust is critical. To address this, we propose an explainable AI framework featuring the Multi-Scale Decomposition Mixing Autoencoder (MDM-AE). Its novel ‘top-down hierarchical mixing strategy’ is physics-informed, fusing the data’s physical context to learn a geometrically superior latent space from its intrinsic manifold. To ensure transparency, we introduce and improve upon ‘decoder-weighted Grad-CAM,’ an interpretability method that provides scientifically consistent visual explanations for the model’s feature extraction process. Evaluated on public spectral datasets, MDM-AE demonstrates significantly superior unsupervised clustering quality over baselines like PCA and standard autoencoders. Our interpretability analysis confirms that this performance stems from an adaptive, chemically sound feature selection strategy. This framework surpasses classical model performance while providing the transparency required for trustworthy scientific discovery from unlabelled sequential data.

General-Purpose Graphics Processing Units (GPGPUs) are crucial for parallel computing in artificial intelligence and big data with their performance heavily relying on efficient warp scheduling. Traditional schedulers, such as Round-Robin (RR) and Greedy-Then-Oldest (GTO), employ static strategies that struggle with adapting to diverse workloads, causing performance disparities across different applications. Prior work has focused on aspects like critical warps and memory access locality but has often overlooked the arithmetic intensity of workloads. Drawing inspiration from the Roofline model and recognizing that different workloads exhibit distinct computational intensities, we propose an Arithmetic Intensity based CacheConscious Adaptive Warp Scheduler (AICAWS). It operates by first analyzing the kernel's static arithmetic intensity through compiler, which serves as a baseline for the hardware. Subsequently, during warp execution, AICAWS dynamically monitors the warp's execution progress, analyzes its runtime arithmetic intensity, and adjusts warp scheduling strategies based on this. Furthermore, AICAWS considers cache locality during warp execution, enabling fine-grained classification of warps based on this. This synergistic mechanism enables AICAWS to effectively hide long-latency memory access operations. Evaluations on diverse benchmarks demonstrate that AICAWS achieves an average performance improvement of 26.3% compared to the baseline scheduler, with a peak improvement of 77.9%.

Based on existing research, we hypothesize that the positioning performance of Time Difference of Arrival (TDOA) systems may exhibit specific spatial distribution characteristics. Through two approaches—performance metric function construction and derivation and geometric intersection analysis—this study reveals the spatial distribution characteristics of positioning performance in minimal 3D-TDOA systems. We systematically establish and refine the theoretical framework for this research. The study confirms the hypothesis, demonstrating that positioning accuracy (PA) is a function of input parameter precision, target location, and sensor placement. Furthermore, we propose and develop core concepts, fundamental theorems, and key properties, including PA boundary (PAB), algorithm shadow zones (regions with poor localization performance), and tangential conditions. This theoretical framework exhibits strong generalizability and applicability. Simulations and experiments validate the theoretical findings. Additionally, we explore the practical applications of this research in target location optimization and sensor deployment optimization. This study provides new theoretical insights for system optimization and performance evaluation, advancing the field of positioning theory.

In this paper, we conducted a group of evaluations on the SpMV kernel with sequential implementation to investigate cache performance on single-core platforms. We verified a similar pattern inside a suite of sparse matrices covering various domains, which makes cache hit rate extraordinary inspiring in a sequential environment. This implicit regularity drove us to propose a cache space splitting approach, aiming at a better locality in dense vector accessing and utilization of large cache capacity in modern processors. Finally, we explored the design space of cache on Matrix 3000 GPDSP and proposed a group of cache parameters, based on our experimental results.