Papers

This paper investigates pinching-antenna systems with finite antenna movement speed, under which conventional antenna placement is subject to non-negligible repositioning delay, resulting in a fundamental tradeoff between channel quality and effective transmission time. In this context, antenna roaming is proposed as a novel operation mode, in which the antenna moves continuously along the waveguide while simultaneously serving users. By incorporating communication and antenna movement into the same transmission cycle, a unified cycle-duration framework is established to facilitate a fair comparison between antenna placement and antenna roaming. The sum-rate difference is then analytically derived, indicating that antenna roaming avoids dedicated positioning overhead with a loss in channel quality due to spatial averaging. The corresponding sum rate maximization problems are formulated for the two operation modes. The continuous optimization problems are transformed into finite-state sequential decision problems and solved via dynamic programming (DP) based algorithms. For antenna roaming, the optimal unconstrained service-interval partition is analytically characterized, with each antenna position assigned to the user achieving the highest instantaneous rate. Simulation results validate the theoretical analysis, demonstrate the performance advantage of antenna roaming over antenna placement, especially when positioning overhead is significant, and confirm the effectiveness of the DP based solutions in improving the achievable sum rate.

A recent construction of unrolled networks for graph-based image restoration forms a system matrix from a graph-Laplacian denoiser through a truncated Taylor expansion, then inverts it with a fixed number of conjugate-gradient steps, with the coefficients of both stages learned. This paper shows the resulting map is a polynomial in the denoising operator, of degree at most the product of the two truncation orders, for every setting of those coefficients and therefore at every point of training: the learned steps select an element of a Krylov subspace they cannot enlarge. At the orders used in practice the reachable set is moreover a measure-zero subset of the polynomial class of the network's own degree budget, so the composition constrains the hypothesis space rather than enlarging it. At the standard initialization the realized spectral response is obtained in closed form, exceeding the intended response throughout the interior of the spectrum and approaching a nonzero floor. A lower bound on the operator's condition number, internal to the graph construction rather than to image content, then places the order required for a prescribed accuracy well above the order used in practice. The confining class is precisely the spectral graph filters for which a direct, convex parameterization has long been available.

Monostatic localization of multiple point targets is studied for a holographic aperture operated through wavenumber-domain modes. A specular-point condition delimits the validity of the spectral model as a near-field approximation. A single snapshot observes a projection of dimension at most the target count times the polarization components, while invertible coding recovers the full channel and places the decoded data on the difference lattice of transmit and receive wavenumbers. Rank conditions settle identifiability, and the Fisher matrix reduces to a covariance over the lattice, dictating a nested mode selection that attains full-aperture resolution with only tens of RF chains.

Echosounders, or high-frequency active sonar systems, have become standard tools for quantifying and mapping the distribution of marine organisms in fisheries or ecological surveys. Conventional echosounder data analysis often relies on human annotation of echograms, which are sonar imagery formed by echo intensity. Over the past decade, in parallel with the exponentially growing volume of echosounder data, there has been a corresponding increase in the development of machine learning (ML) methods that operate primarily on echograms as images. However, echograms are not simply images: they are associated with specific spatiotemporal coordinates that are essential for alignment with survey events, human annotations, and other oceanographic datasets. We present a generalizable two-stage workflow for constructing analysis-ready datasets for ML development tailored for echograms from transect-based surveys, in which (1) acoustic data are partitioned according to transect designation, and (2) masks are created from annotations referencing user-defined uniform spatiotemporal echo data grid. Importantly, ancillary information, such as geospatial coordinates and oceanographic measurements, is propagated across processing stages to preserve the essential contextual information for downstream analyses. We demonstrate the scalability of our workflow implementation based on two open-source software libraries, Echopype and Echoregions, using two example fisheries survey datasets. We additionally provide an executable tutorial that guides readers through the computational implementation of this workflow. Together, these elements provide a scalable and generalizable framework for creating analysis-ready echosounder datasets for ML applications.

Meditation has been associated with a range of cognitive and physiological benefits. However, the underlying neural mechanisms of different meditation practices are not yet fully understood. This study investigates whether electroencephalogram (EEG) signals can be used to characterize and distinguish two commonly practiced meditation techniques, namely Samatha and Vipassana, which involve distinct cognitive processes, with Samatha emphasizing sustained concentration and Vipassana emphasizing mindful observation and insight. By extracting features such as band power, coherence and wavelet entropy from EEG signals recorded during pre-meditation resting, Samatha, and Vipassana states, we provide an assessment of how spectral power, signal complexity, and functional connectivity differentiate these practices. Preliminary results from $12$ experienced meditators indicate condition-dependent EEG differences, with the most pronounced effects observed in the delta frequency band. These results demonstrate the utility of EEG-based metrics for the objective characterization of meditation and advance our comprehension of the neural processes underlying different contemplative practices.

Thoracic pathologies rarely occur in isolation, yet standard multi-label classifiers rely on shared global descriptors, discarding where findings lie and how they co-occur. We propose \textbf{C$\mathbf{^2}$A} (Co-occurrence Aware Class Attention), a classification head that explicitly couples spatial evidence with clinical priors. First, C$^2$A casts pooling as an expectation over learned per-class spatial attention maps, yielding localized descriptors for each disease. Second, it couples these descriptors via a learnable graph warm-started from empirical label co-occurrence. A single residual message-passing step shares evidence among related findings, proving to be a bounded perturbation of the identity where co-occurrence enters each logit through an explicit bilinear interaction. On CheXpert, C$^2$A achieves a superior $0.895$ macro-mean AUROC, outperforming advanced context-gating baselines. Crucially, gains concentrate on highly co-occurrent classes with ambiguous spatial evidence (rescuing Atelectasis by $+1.5$ over GCG), demonstrating the prior's regularizing effect with a negligible overhead of one linear projection and a $C\!\times\!C$ edge matrix.

Retinal fundus images frequently exhibit multiple co-occurring pathologies, yet standard deep learning classifiers apply static, identical computation to every image regardless of the underlying disease distribution. We propose a novel architecture that resolves this via sparse conditional computation, pairing a Guided Context Gating (GCG) spatial attention front-end with a sparsely-routed Mixture-of-Experts (MoE) block operating over feature tokens. Crucially, this routing yields an interpretable, data-driven decomposition. Expert allocation is significantly disease-dependent (p < 0.001), with the healthy Normal state and morphologically distinct pathologies (e.g., ERM, AMD) isolating to dedicated experts. On a five-class, patient-disjoint 5-fold cross-validation benchmark, our model achieves 0.912 +/- 0.008 macro AUC and 0.653 +/- 0.014 macro F1. Furthermore, Grad-CAM++ and post-MoE t-SNE visualizations confirm that expert routing aligns with localized lesions and geometrically maps co-occurring cases between their constituent clusters, positioning sparse MoE as an interpretable approach to multi-disease retinal screening.

Synthetic radio-frequency (RF) sensing data are widely used to augment wireless sensing tasks, yet their measurement consistency with real data is rarely evaluated under matched acquisition conditions. This paper identifies a measurement-consistency failure mode: synthetic samples may pass task-facing checks while deviating from the measurement behavior of real samples collected and processed by the same sensing pipeline, potentially introducing synthetic shortcuts and biasing downstream model selection. We propose RFCheck, a calibrated measurement audit that uses held-out real data from the same acquisition and preprocessing pipeline as the reference. RFCheck calibrates representation-specific tests on real samples and flags synthetic samples whose responses exceed the calibrated real-data range. We use the audit for candidate screening and residual repair. We validate RFCheck primarily on Wi-Fi channel state information (CSI), where the audit examines delay-domain and local frequency-domain structures. Experiments show that aggregate statistics and label-based screening can miss measurement failures detected by RFCheck. Under the same label acceptance rule, low-risk and high-risk synthetic candidates exhibit different downstream behavior. A repair reference reduces the flagged ratio to 10.83% while preserving mean task performance. In a held-out proposal study, correction followed by calibrated selection produces a class-balanced set with no flagged samples under a fixed training budget. We further apply the same calibration principle to frequency-modulated continuous-wave (FMCW) millimeter-wave radar gesture sensing. The results show that synthetic RF sensing data can violate measurement consistency even when conventional task checks are satisfied, motivating measurement-aware diagnosis and mitigation before augmentation.

Next-generation data-channel applications of polar codes demand decoding algorithms with high throughput and low latency. The recently proposed speculative successive cancellation (Spec-SC) decoding reduces average decoding latency by speculatively executing the right branch of successive cancellation (SC) decoding in parallel with the left branch, verifying the result once the g-function is computed. In this paper, we propose new acceptance criteria based on Hamming distance and ellipsoidal distance that improve acceptance rates, enabling greater speed-ups for lower-rate nodes. We further show that combining code membership testing with speculative decoding accelerates high-rate nodes as well. Numerical results confirm that both approaches outperform fast simplified SC (FSSC) decoding and Spec-SC with the original acceptance condition in terms latency by a wide margin, with virtually identical error-rate performance.

Lane detection requires recovering thin, elongated, and frequently occluded lane structures under challenging driving conditions. While anchor-based detectors provide efficient candidate generation, their performance is limited by two coupled issues: backbone features often lose structural continuity along partially visible lanes, and classification confidence may decouple from line-level localization quality, allowing inaccurate anchors to persist before non-maximum suppression (NMS). We propose a structure-enhanced and quality-aware framework that improves lane representation and dynamic-anchor scoring while preserving the inference pipeline of the Anchor Decomposition Network (ADNet). Specifically, a Gated Horizontal-Vertical Token (GHVT) module enhances mid- and high-level backbone features via lightweight directional token interactions with a learnable residual gate. In parallel, Line-Quality-Aware Dynamic Anchor Scoring (LQAS) calibrates existing classification logits using quality supervision, hard-negative suppression, and pairwise ranking without adding inference branches. On the VIL-100 dataset, our method improves ADNet-R34 from 89.97 to 91.28 in F1 score at the 0.5 intersection-over-union threshold (F1@50), reducing both false positives and false negatives. Additional experiments on CULane and TuSimple datasets, extensive ablations, score-distribution diagnostics, and runtime analysis confirm complementary structural and ranking improvements with minimal computational overhead.

Optimization theory is a widely used tool for intelligent decision-making. While classical optimization deals with fixed, time-invariant objective functions, many modern applications operate in dynamic environments where data arrive sequentially, and the learning objective evolves over time, often under decentralized data and communication constraints. Motivated by these trends, we study decentralized optimization from streaming data through a structured time-varying formulation in which the global objective is a temporally weighted average of losses observed across the network. We analyze multi-iteration decentralized first-order methods, including decentralized gradient descent. For strongly convex and smooth losses, we develop guarantees for the Euclidean-norm tracking error through a contraction-mapping viewpoint. The resulting bounds decompose the tracking error into a fixed-point tracking component and a bias term induced by decentralization and data heterogeneity. We specialize our analysis to uniform and exponentially discounted weights, as well as their finite-memory windowed counterparts. The bounds explicitly characterize the roles of the temporal weighting rule, per-step iteration budget, step size, and network connectivity. Uniform weighting yields a vanishing fixed-point tracking contribution of order $\mathcal O(1/t)$, whereas discounted and windowed strategies generally induce non-vanishing tracking floors governed by the discount factor and effective memory, respectively. In all cases, decentralization induces an additional non-zero bias floor under a constant step size. Numerical experiments illustrate the predicted trends.

Ensemble decoding is a promising technique for ultra-reliable low-latency communication, as it trades hardware parallelism for decoding latency by running M diverse belief propagation (BP) decoders in parallel. Conventional ensembles select their output from the candidates. Hence, decoding fails whenever no member finds the correct codeword. In this work, we show that shallow, diverse BP members are poor correctors, but excellent sorters. Based on this observation, we propose a soft ensemble-combining stage (SECS) that combines extrinsic messages after only a few iterations, yielding a reliability ordering whose most reliable positions are nearly error-free. A subsequent re-encoding stage (like ordered-statistics decoding (OSD)) converts this ordering into a near-maximum likelihood (ML) candidate codeword, at a fraction of the BP latency of a fully converged ensemble. We demonstrate the proposed scheme on three short codes: a (63,30) Bose-Ray-Chaudhuri-Hocquenghem code, an overcomplete PG(2,8) code, and a search-designed (105,53) cyclic code. In all cases, the SECS with OSD post-processing closes most of the gap to ML decoding, while significantly reducing the number of required BP iterations.

Belief-propagation (BP) decoding of short and moderate-length low-density parity-check (LDPC) codes is limited by finite-length graph effects: a single decoder trajectory can become trapped or oscillatory even when an alternative trajectory would decode the received word. Existing ensemble-BP decoders create the required diversity through multiple parity-check matrices, automorphisms, modified schedules, subcodes, or altered update rules. We introduce row-boosted ensemble (RBE) decoding as a minimal decoder-side diversity mechanism: all ensemble members share the same parity-check matrix and the same BP kernel, and differ only in a small set of parity-check rows whose outgoing messages are boosted. On the 5G~NR BG1 \((144,96)\) code, RBE with \(32\) members lowers the frame-error rate of BP with \(20\) iterations (BP-20) from \(1.6\times10^{-2}\) to \(1.6\times10^{-3}\) at \(E_\mathrm{b}/N_0=4.0\,\mathrm{dB}\), outperforming saturated-min-sum and affine subcode ensembles of equal size. Increasing the ensemble size yields additional gains, indicating that RBE provides a scalable performance-complexity tradeoff. The gains transfer across 5G~NR block lengths and rates, to non-5G short LDPC codes, and across flooding and layered schedules.

Multi-camera wireless perception requires a base station (BS) to maintain timely and reliable latent beliefs from distributed cameras under limited uplink resources. Conventional Age-of-Information (AoI) measures the age of the latest received update but not task-relevant latent content. We introduce Age-of-Latent (AoL) to quantify the freshness of each camera's latest decoded latent representation. A finite temporal window of integration (TWI) determines the task-commitment time and available uplink slots, creating a tradeoff among update opportunities, prediction duration, commit-time AoL, prediction reliability, and accumulated resource cost. Within this horizon, redundant or overlapping views enable correlated prediction, reducing reliance on the highest-cost communication and encoding configuration to maintain BS-side latent beliefs. We formulate a joint AoL-resource minimization problem coupling task-level TWI selection with slot-level encoder selection, scheduling, and NOMA power allocation under prediction-reliability constraints. We propose correlation-aware latent prediction for AoL minimization (CoLA), which uses Lyapunov optimization for task-level TWI selection based on AoL-resource cost and prediction uncertainty, and proximal policy optimization for slot-level resource control. Results on a warehouse multi-camera RF dataset show that CoLA adapts the TWI to camera-update availability and achieves the most favorable AoL-resource tradeoff among the benchmarks while maintaining prediction reliability, particularly under prolonged and severe camera outages.

State-of-Polarization sensing with coherent transponders enables wide-area geophysical monitoring over existing submarine cables, but its performance is limited by polarization noise from both the fiber and terminal hardware. This work investigates how terminal noise affects polarization rotation estimates derived from receiver equalizer coefficients and how it can obscure stochastic polarization drift used for sensing. We analyze Jones-matrix time series from two deployed receivers in the Sparkle Mediterranean link MedNautilus (approximately 2000 km and 450 km) and compare them with a laboratory back-to-back reference. Power Spectral Density (PSD) analysis reveals a low-frequency random-walk regime and a high-frequency white-noise floor, separated by a link-dependent corner frequency. The rotation innovation variance increases with link length, while the longest field link also shows elevated white noise consistent with accumulated amplifier and terminal contributions. Additionally, harmonic spectral components are observed, indicating a transponder-related artifact that should be considered in practical sensing applications.

In this paper, we investigate the angle-of-arrival (AoA) estimation problem for wireless sensing systems equipped with movable antennas (MA). To achieve high estimation performance and accuracy, we formulate a joint optimization problem integrating the sidelobes of steering vector correlation (SVC) and the Cramér-Rao bound (CRB). We first mathematically transform the SVC and the CRB into tractable objective functions. Specifically, we introduce a proxy variable and apply a discrete grid search strategy to overcome the intractability of optimizing the SVC with unknown target angles. Concurrently, we derive a generalized lower bound for the CRB, which yields a scalar function of the MA positions. Guided by the transformed objective, we propose a successive convex approximation-based position optimization algorithm. The proposed algorithm handles the non-convex terms by employing first order Taylor expansions within a defined trust region, which allows the MA positions to be updated incrementally in each iteration. Simulation results demonstrate that the proposed algorithm achieves superior AoA estimation performance.

Wireless enhancement is critical for next-generation mobile communication systems to realize seamless connectivity, yet traditional network expansion strategies are becoming economically unsustainable. Reconfigurable intelligent surfaces (RISs) provide a promising alternative by improving signal utilization. However, high hardware and deployment costs of advanced RISs limit their large-scale application. Here, we democratize this technology with OpenRIS, an open-source and low-cost platform composed of Lego-like meta-bricks. With digital-twin assistance, these meta-bricks can be flexibly assembled into arbitrary shapes to achieve customized, mass-deployable wireless enhancement without extra power. Experiments and full-wave simulations verify that the discretized OpenRIS achieves consistent performance with the continuous RIS. We further develop a dual-user wireless transmission system and a three-dimensional coverage measurement system to showcase the versatile applicability of OpenRIS in wireless enhancements. As a plug-and-play solution, OpenRIS accelerates the translation of RIS theory into practice and is poised to integrate into infrastructure, reshaping the future wireless world as steel and concrete shape modern cities.

Model learning by an eavesdropper is treated as an estimation problem in a federated environment. The Fisher Information Matrix for the eavesdropper's estimation problem is driven to singularity through a signaling design; this ensures that the eavesdropper cannot learn the model. Herein, the innovation of prior designs is that model shifts are designed to maximize the difference in the model learned by Eve and the central server while satisfying a transmission power constraint for the agents. Two shift schemes are provided. MaxModShift outperforms a prior ModShift design while requiring lesser transmission power. Compared to a noise injection scheme, MaxModShift performs better while requiring a lower bandwidth secret channel and a reduced average power consumption.

Learning-based wireless localizers often fail to utilize geometric information about the propagation environment, limiting their ability to exploit non-line-of-sight (NLoS) propagation and generalize across scenes. To bridge this gap, we propose GLocFM, a Geometry-aware Localization Foundation Model, which jointly exploits WiFi measurements and scene geometry represented as a 3D point cloud. We formulate localization as a maximum-likelihood (ML) estimation problem, where the goal is to find a transmitter position that maximizes the likelihood of the wireless observations conditioned on the scene geometry. The likelihood of a candidate transmitter position is calculated by a learned scoring function that matches the observed delay–angle-of-arrival (AoA) spectrum against the spectrum predicted for that candidate. A hierarchical scene encoder extracts propagation-relevant features to produce geometric priors for LoS and one-bounce reflection paths. For scenarios with imperfect synchronization, we further introduce a time-of-flight (ToF)-robust GLocFM model to handle unknown ToF offsets. GLocFM is trained on a multi-modal synthetic indoor localization dataset comprising 221 diverse scenes whose associated wireless signals are generated using Sionna RT. On both synthetic and the NeRF$^{2}$ dataset based on real measurements, GLocFM reduces mean 3D localization error relative to one of the state-of-the-art localization baselines by 49.5% and 48.8%, respectively. Ablations across different number of receiver, bandwidths, and array sizes further demonstrate the effectiveness and robustness of the proposed framework.

Detection of atrial fibrillation (AF) from photoplethysmogram (PPG) is highly sensitive to artifacts, yet their effect on uncertainty of different AF detectors remains poorly understood. This work aims to quantify how different PPG artifact types affect the uncertainty of AF detectors. Two machine learning approaches to AF detection were explored: one using 25-s PPG signals as input ($\mathit{D}_r$) and another using AF-related rhythm irregularity features ($\mathit{D}_f$). The detectors were trained on wrist PPG signals acquired during cardiac rehabilitation and then systematically evaluated on 260,000 PPG signals containing controlled artifact types and durations. Uncertainty was quantified using a threshold-based error rate, conformal prediction, and Monte Carlo dropout. Using artifact-free PPG signals, $\mathit{D}_f$ outperforms $\mathit{D}_r$ with sensitivity/specificity of 0.94/0.94 versus 0.92/0.86. Relative to artifact-free performance, sensitivity/specificity drops by 0.52/0.03, 0.25/0.03, 0.16/0.02, and 0.07/0.02 using $\mathit{D}_r$ for 12-s artifacts of device displacement, forearm motion, hand motion, and poor contact respectively. For $\mathit{D}_f$, the respective drops are 0.31/0.07, 0.19/0.11, 0.14/0.15, and 0.13/0.21 for the same artifacts. $\mathit{D}_f$ is more robust to short artifacts but exhibits increasing uncertainty with longer artifact durations, whereas $\mathit{D}_r$ shows an abrupt performance drop when artifacts occur but is less sensitive to artifact duration. Applying conformal prediction with 90$\%$ coverage reduces the false-positive rate by up to 12$\%$ for $\mathit{D}_r$ and up to 64$\%$ for $\mathit{D}_f$. Artifact type and duration have detector-specific effects on AF detection uncertainty. Device displacement causes the largest increase in uncertainty.

Low Earth orbit (LEO) satellite networks are envisioned as a promising solution for providing ubiquitous connectivity and narrowing the digital divide. The extensive footprint of LEO satellite constellations enables broad coverage, resulting in spatially non-uniform traffic demand across the serviced areas. Meanwhile, stringent on-board power constraints make power-intensive transmission architectures less attractive and motivate energy-efficient transmission strategies that effectively exploit scarce satellite network resources. To this end, this paper proposes a cooperative transmission framework that jointly accounts for non-uniform traffic demand and network-wide power consumption. Each LEO satellite integrates hybrid precoding (HPC), radio frequency (RF) chain activation, and hardware quantization, while user-equipment (UE)-centric satellite clusters are organized using statistical channel state information (sCSI) and traffic demands. A framework for joint optimization of cooperative transmission architecture and resource allocation is designed to maximize demand-aware energy efficiency (EE), resulting in a mixed-integer nonlinear program (MINLP) for which finding a globally optimal solution is generally intractable. Accordingly, a two-stage algorithm is developed under a distributed linear precoding structure, in which a modified cross-entropy (CE) method searches over discrete variables, while fractional programming is employed for transmit power allocation. Numerical results indicate that the proposed framework outperforms benchmark schemes while accounting for traffic demands and EE.

Cardiologists interpret electrocardiograms by localizing waveform components, measuring rhythm and interval patterns, and translating these structured observations into diagnostic evidence. Whether this expert reading process can serve as an effective prior for ECG agents remains unclear. To address this question, we introduce LuminaECG, a clinically structured ECG reasoning framework that reformulates ECG interpretation as measurement-grounded visual reading. ECG signals are rendered on standard electrocardiographic grid paper to preserve the spatial and scale cues used in clinical reading. P-wave, QRS-complex, and T-wave boundaries are explicitly delineated, and color-coded segmentation decomposes the waveform into discrete visual measurement primitives. A general 2B vision-language backbone is then trained with low-rank supervised fine-tuning to associate these primitives with diagnostic reasoning, without architectural modification. Across open, proprietary, and ECG-specialist zero-shot baselines, LuminaECG improves both waveform measurement and diagnostic recovery. It reaches a clinically meaningful reader tier on the CODE-test benchmark, transfers across geographically diverse ECG datasets without retraining, and generates reports whose structure contains an emergent prognostic signal. These findings suggest that effective ECG agents require not only larger models, but supervision that preserves the alignment between measurable waveform evidence and clinical knowledge.

Semantic feedback can repair a noisy first decode without retransmitting the whole message, but current end-to-end evaluations hide where a feedback design fails. An oracle that sees both the source-side state and receiver observation may describe a useful correction with few bits, while the receiver, which sees only its own observation, may not know which correction to request. We formulate these as distinct direct and receiver-constrained coding problems, connect repair distortion to retained local utility and prove a random-access separation: one direct bit is exact although receiver utility vanishes with the number of source-conditioned alternatives. A synthetic ladder verifies the exact law. In paired populations of eight Reuters text-JSCC codecs, the trained recurrent population retains over three times more direct 16-bit repair value than the trained DeepSC population despite comparable full-repair opportunity. Nevertheless, an entropy-valid receiver audit across all eight recurrent checkpoints does not resolve against repetition, source-only or shuffled-index controls. The results show that compressible repair is an upper bound, not evidence of deployable feedback, and motivate reporting both rate curves under matched resources.

Partial-Fourier measurement underlies computational imaging, and its reconstructions increasingly come from iterative or learned solvers whose recovery guarantees are conditional on a signal model, a sampling law, and solver accuracy. None of those guarantees transfers to a particular committed output: it can satisfy every acquired coefficient while remaining badly wrong in the unmeasured nullspace. Native Fourier holdout does not repair this, since a spectrally concentrated error is missed unless its bin is drawn, so uniform worst-case risk needs a number of hidden bins proportional to the dimension. We propose a post-commit acceptance test: secret finite-phase masks turn one fixed Fourier readout into dense randomized residual checks, and a median-of-means threshold handles calibrated readout noise. We prove that errors above $E_{\rm rej}$ are rejected and errors below $E_{\rm acc}$ are accepted with confidence $1-\delta$ once $q\ge C(1+N\sigma^2/E_{\rm rej})^2\log(1/\delta)$, explicitly exposing the one-bin resolution floor $N\sigma^2/\sqrt q$; in the zero-noise limit, QPSK gives $4^{-q}$ soundness versus $97.9\%$ minimax false acceptance for a worst-case ordinary-bin error. At a fixed noise-to-threshold ratio, measured screening false acceptance fell from about $21\%$ at 16 checks to $5\%$ at 64; a data-consistent learned nullspace failure with NMSE $0.200$ was rejected by $96.9\%$ of challenge banks while 16 ordinary bins missed it $98.9\%$ of the time; and error statistics at 16 checks stayed stable from 128 through $65{,}536$ dimensions. The test supports solver acceptance, model selection, and stopping certification, with a certified fresh-bank error interval.

Hyperlocal meteorological sensing is essential for accurate solar photovoltaic forecasting, yet professional-grade meteorological stations require investments easily exceeding 1000~USD per node, making distributed deployments economically inaccessible. This work presents a modular internet of things (IoT) device based on the ESP32 microcontroller integrating temperature, humidity, luminosity, and solar irradiance sensors in an IP68-rated enclosure at a total hardware costs of about \$65~USD when components are sourced in Germany. A hybrid architecture decouples external model training, performed on a conventional computer using the software Python and the open-source library TensorFlow, from autonomous 24-hour solar voltage forecasting executed on-device via a three-layer feedforward network with 3{,}011 parameters (11.8\,KB). The network is trained offline on site-collected data and deployed on the microcontroller as static weight matrices without cloud connectivity. An on-device incremental gradient descent mechanism enables continuous model adaptation after deployment without external retraining. The system was evaluated through two field deployments: a short period of hardware and firmware validation in Ulm, Germany, and a 115-day deployment in Zapopan, Mexico, comprising 84~days of training and 31~days of autonomous operation with zero missing records. Over a clean 28-day daytime window, the embedded model attained a coefficient of determination of 0.9165 and a mean absolute error of 0.2975~V (4.65\% of the operational range), outperforming a climatology baseline (skill score 0.64) while not surpassing a 24-hour persistence baseline. A frozen-weight ablation confirms that the on-device update mechanism yields a small but statistically robust accuracy gain ($p = 0.001$), demonstrating that autonomous incremental learning is feasible on low-cost hardware without cloud connectivity.

Foundation models are emerging as a transformative paradigm for AI-native sixth-generation (6G) wireless networks by enabling scalable, transferable, and data-efficient intelligence across diverse communication tasks. Unlike conventional deep learning models that are trained for individual applications, wireless foundation models (WFMs) learn generalized representations from large-scale heterogeneous wireless data and can be efficiently adapted to communication, sensing, localization, and network optimization tasks with minimal task-specific supervision. Despite rapid progress, current research remains fragmented across architectures, training paradigms, and application domains, with no unified survey dedicated to the design, learning, and deployment of WFMs. This survey presents a comprehensive and unified review of wireless foundation models. We first establish the fundamental concepts of WFMs and introduce a taxonomy that organizes the field according to model architectures, pre-training paradigms, and applications. We then review representative architectures, self-supervised pre-training strategies, parameter-efficient adaptation methods, datasets, benchmarks, and evaluation methodologies, highlighting their roles in enabling transferable wireless intelligence. Furthermore, we examine emerging applications spanning physical-layer signal processing, network intelligence, and cross-layer optimization, and discuss the key challenges of data availability, generalization, interpretability, efficient edge deployment, and standardization. Finally, we outline future research directions toward scalable, trustworthy, and general-purpose wireless intelligence for AI-native 6G networks. This survey provides a comprehensive reference for researchers and practitioners developing next-generation intelligent wireless systems.

Cooperative multi-unmanned aerial vehicle (UAV) reconnaissance is often hindered by intermittent air-to-air communications where link dropouts lead to uncoordinated exploration and redundant mapping. Existing approaches rely on explicit exchange of high-dimensional spatial data or raw observations, incurring significant overhead, and often revert to reactive individual exploration during outages. This paper proposes a memory-augmented framework in which each UAV maintains a structured latent state decomposed into map, task, and memory components. During dropout, a generative predictor conditioned on the memory state infers substitute peer messages in the latent space, making the estimation task more tractable and directly aligned with the cooperative objective. The framework is trained end-to-end under the centralized training with decentralized execution paradigm. Simulation results demonstrate that the proposed framework closely matches the performance of a fully connected swarm, while remaining robust across a wide range of link failure conditions.

State departments of transportation (DOTs) in the United States increasingly rely on statewide continuously operating reference station (CORS) networks to support high-precision Global Navigation Satellite System (GNSS)-based positioning and timing for intelligent transportation systems. These networks also provide continuous observations that can support regional GNSS integrity monitoring. This study develops and demonstrates a framework that treats a statewide CORS network as a spatially distributed sensor system for identifying unintentional (environmental) and intentional (cyber) interference when GNSS measurements deviate from expected spatial patterns. We develop a graph-based Network Consistency Framework (NCF) that evaluates each station against its spatial neighborhood using four metrics: neighborhood residual, spatial gradient, residual, and graph smoothness. These metrics are combined into a Network Consistency Index (NCI). The framework is demonstrated using two consecutive days of four-constellation observations from 50 stations in the Alabama DOT-maintained CORS network, using changes in vertical total electron content ({\Delta}VTEC) and the Rate of TEC Index (ROTI) as spatially coherent observables. The framework quantified network-wide spatial consistency and identified localized anomalies. Detected anomalies indicate stations whose observations deviated from the surrounding regional network, signaling potential integrity issues. Determining whether anomalies result from receiver faults, localized interference, spoofing, or other causes requires further investigation. This study introduces statewide CORS networks as regional GNSS integrity observatories and presents the NCF and NCI for graph-based spatial integrity monitoring. Transportation agencies can implement the framework using existing CORS observations to monitor network integrity and identify localized anomalies.

Koopman theory offers a linear-operator view of nonlinear sequence dynamics by lifting observations into a space where evolution is governed by a linear time-invariant Koopman operator. While the Koopman operator provides a linear representation of nonlinear dynamics, it is generally infinite dimensional and defined under time-invariant assumptions. To model non-stationary signals with frequency-dependent behavior, a frequency-varying extension is required. In recent years, deep learning has been increasingly employed to exploit its powerful function-approximation ability for learning the Koopman operator. In this study, we propose a novel approach called neural decomposition Koopman (NDKoop), an end-to-end architecture that integrates a learnable signal decomposition module with both frequency-independent and frequency-dependent Koopman based networks for sequence forecasting. To the best of our knowledge, this is the first work to jointly realize end-to end Koopman modeling and signal decomposition within a unified neural framework. We demonstrate that decomposing a signal into a frequency-independent trend component and a frequency-dependent periodic component, each governed by a corresponding Koopman operator, improves prediction accuracy when perfect linearization is unattainable. Numerical experiments across several forecasting benchmarks indicate that the proposed NDKoop provides strong performance.

Connection graphs (CGs) extend classical graphs by associating vector-valued signals to nodes and orthogonal transport maps across edges, making them a natural model for synchronization and manifold-based signal processing. Despite their growing use, learning CGs directly from observations remains challenging because the network topology and the underlying geometric structure are coupled through non-Euclidean orthogonality constraints. In this work, we address this inverse problem by learning a consistent connection graph from noisy vector-valued signals. Exploiting the spectral characterization of consistent CGs, we formulate a structured learning problem that jointly estimates a denoised signal, the graph topology, and node-wise local reference frames. The proposed formulation couples the spectrum of the learned connection Laplacian to that of an underlying combinatorial Laplacian, enabling explicit spectral and topological priors while guaranteeing a nontrivial global-section space. We develop Structured Connection Graph Learning (SCGL), a block-coordinate algorithm that combines closed-form updates, manifold projections, and spectral constraints, and converges to stationary points of the resulting nonconvex problem. Numerical experiments show that SCGL improves topology and geometry recovery over competing approaches, while also yielding effective denoising and signal-compression bases.