Papers

Orbital angular momentum (OAM) multiplexing can increase the capacity of free-space optical (FSO) communications, but its detection performance is strongly affected by impairments such as atmospheric turbulence, transmitter pointing errors, and photodetection noise. The diffractive deep neural network (D$^2$NN) can be used as an all-optical front end to mitigate turbulence-induced distortions before detection. However, existing D$^2$NN compensation schemes are not specifically optimized for communication detection. In this paper, we propose a supervised contrastive inspired D$^2$NN (SCI-D$^2$NN) framework for improving the detection performance of OAM-multiplexed FSO communications under these impairments. The proposed framework introduces two training branches: a projection branch that maps the optical field to low-dimensional decision domain samples, and a label branch that provides supervised labels to impose a separation constraint among decision domain samples. In addition, we characterize complex-amplitude crosstalk to obtain the receiver observation vector and formulate two detection schemes, namely single-port profile-likelihood detection and joint maximum-likelihood (ML) detection. We further design two SCI-D$^2$NN training losses called Bhattacharyya distance (BD) based loss and the ML based loss to improve decision domain separability and mitigate detection-performance degradation. Numerical results show that SCI-D$^2$NN achieves more than a 3-dB improvement in bit error rate (BER) over the conventional D$^2$NN baseline in most transmit-power regions. The BD based loss gives the lowest BER under different system parameters and provides more than a 10-dB BER improvement over the baseline in the high transmit power region.

Generative music systems are increasingly presented as tools that democratize music creation, yet their practical suitability for musicians remains underexplored. Prior work includes openness-focused evaluation frameworks, such as MusGO (Music-Generative Open AI), as well as qualitative studies of musicians' experiences with generative systems. However, these approaches do not support systematic comparison or early-stage discovery of models for creative use. Motivated by such limitations, we introduce MusGU+, a musician-centered evaluation framework organized around three dimensions: Adaptability, Usability, and Controllability. Together, these capture whether a model can be feasibly trained or fine-tuned on personal data, integrated into real-world music workflows, and controlled in musically meaningful ways. We evaluate 10 representative generative music systems and present an interactive discovery tool that enables musicians to explore and filter models according to these criteria. While MusGO remains valuable for promoting responsible research practices, MusGU+ supports informed selection and practical adoption of generative systems by musicians.

Forward modeling of scattering centers of radar targets is critical for advanced information retrieval of Synthetic Aperture Radar (SAR) images. Existing forward modeling approaches rely on meshing the target and computing scattering via ray-tracing techniques, which not only incur high computational cost but also discard the semantic information of target geometry, thereby limiting the interpretability of SAR imagery. To address these issues, this paper proposes a novel forward scattering modeling framework that directly constructs stable, compact, and physically meaningful scatterers from target geometry. This formulation decouples target representation from specific observation configurations, enabling an intrinsic scatterer description across varying viewing angles. Specifically, an improved Random Sample Consensus (RANSAC) scheme is first developed to robustly extract planes, cylinders, and spheres from target point clouds, yielding single-bounce scatterers. Potential multiple-bounce scatterers are then automatically detected by analyzing inter-primitive relations, and further refined through geometric clipping and parameter alignment to produce unique scattering representations compatible with canonical scattering center models. Finally, radar responses and SAR images can be generated under arbitrary observation configurations based on the constructed scatterers. The proposed method is validated through extensive experiments, including comparisons with electromagnetic simulations and measured data. The results show that it can accurately characterize the scattering behaviors of complex targets, providing an efficient, interpretable, and reliable solution for SAR image modeling and understanding.

This article proposes a differentiable physics simulation (DP simulation)-based parameter estimation method for the condition monitoring of power electronic converters. In the proposed method, the time-domain simulation of converter dynamics is embedded into a differentiable computational graph, directly linking device parameters to observed voltage and current trajectories. By formulating differentiable time-stepping operators, the nonlinear dynamics of the converter across different circuit topologies are simulated in a unified, differentiable manner. A dc-dc buck converter is used as a representative case study. Using sparse transient samples from existing sensing channels, the method enables noninvasive parameter estimation without additional sensing hardware. Comprehensive simulation studies are conducted to evaluate the impacts of time-stepping schemes, regularization constraints, and various uncertainty sources on estimation accuracy and robustness. Subsequently, 30 distinct hardware configurations are experimentally tested for validation. The results show that the proposed method can effectively track the relative variations of health-related parameters across the critical components. This DP simulation framework provides a novel perspective for physics-informed machine learning in power electronic applications.

The rapid expansion of large-scale artificial intelligence (AI) data centers is adding substantial, concentrated, and rapidly varying loads to transmission-constrained power systems. Although such load variations are generally regarded as operational challenges, this paper presents an alternative perspective in which the upward load flexibility of AI data centers could be coordinated for transient-stability support. To this end, this paper proposes training-induced load surge (TILS), a fast demand-side strategy that initiates or resumes flexible AI training workloads after fault clearing to increase active-power demand at electrically effective locations. The resulting load increase allows accelerating generators to supply additional electrical power, thereby reducing the accelerating-power imbalance and limiting the first-swing rotor-angle excursion. The underlying mechanism is first clarified in a single-machine infinite-bus (SMIB) system and then evaluated in the IEEE 39-bus system and a large-scale Korean power system. Results across all three systems demonstrate that TILS can increase the transient-stability-constrained generation limit. Larger responses, earlier activation, and siting at buses with a stronger electrical influence on the critical generators provide greater generation-limit increases. These results suggest that the upward load-response capability of AI data centers can provide complementary transient-stability support when sufficient electrical headroom, flexible workloads, and reliable grid-triggered activation are available.

Deploying a new control policy for voltage control in active distribution grids requires evidence that physical limits will be satisfied before the policy is tested on the physical grid. This assessment is difficult for two reasons. First, simulations cannot capture every disturbance, modeling error, and device interaction present in the real grid. Second, historical measurements reflect operation under existing control policies, whereas a new policy may drive the grid into different operating conditions. To address these challenges, we propose Distributionally Robust Conformal Safety Screening (DR-CSS), a policy-agnostic framework for pre-deployment, scenario-by-scenario screening of a new control policy using historical data and a nominal simulator. For each new scenario, the simulator predicts a future voltage trajectory for the whole grid; DR-CSS then constructs a conformal safety interval around this prediction using historical simulation-to-reality errors. The interval is further enlarged to account for closed-loop changes induced by the deployment of the new policy and its interactions with the remaining controllers. To the best of our knowledge, DR-CSS is the first framework in power systems to combine historical data from an existing control policy with an imperfect simulator for pre-deployment safety screening of a new policy. Experiments on the IEEE 33-bus and IEEE 141-bus systems evaluate the deployment of learning-based voltage control policies and show that DR-CSS identifies all unsafe test scenarios. To reduce unnecessary warnings on safe scenarios, we adapt the safety intervals to different operating conditions and gradually introduce new policies with recalibration after each stage. These extensions increase the informational value of the safety screening and support safer deployment decisions in active distribution grids.

This comment revisits Lemma 1 in [1], which plays a central role in the event-triggered stabilization analysis developed therein. We identify technical gaps in the proof of the lemma and provide a corrected argument. In particular, careful treatment of the exponentially decaying term shows that its decay rate must be retained in the resulting convergence estimate. The statement of the original lemma, with the exponential decay rate determined by the minimum of the characteristic decay rate and the decay rate of this term, remains valid.

This paper considers decentralized contingency MPC for multi-agent control under a state-only information pattern, with particular focus on limited sensing and plug-and-play operation. The objective is to retain recursive feasibility, safety, and Lyapunov-type convergence while reducing conservatism in local interaction handling. The framework relies on agent-wise fallback regions (safe sets) in which a feasible contingency maneuver to a safe equilibrium is always available. A novel safe-set update mechanism is introduced that supports less conservative decentralized interaction while preserving the underlying guarantees. This, in turn, enables memory-free local interaction and finite sensing ranges without requiring agents to reconstruct the exact neighbor geometry. The resulting scheme remains fully decentralized and preserves the shared-first-input contingency MPC structure. Theoretical guarantees and simulation results illustrate the effectiveness of the approach in dense multi-agent scenarios.

A measurement principle for determining liquid water content (LWC) in snow is presented that does not require a priori knowledge or separate measurement of snow density. LWC affects the imaginary part of the effective permittivity of snow, which is linked to microwave attenuation along a microstrip transmission line embedded in the snowpack, independent of snow density. By analyzing the attenuation over a range of microwave frequencies, the corresponding LWC can be inferred. Full-wave simulations indicate a sensitivity of approximately 2 dB per % LWC,while experimental results confirm the insensitivity to snow density, highlighting the potential for remote snow characterization.

This letter investigates a task-unified multi-level radio-frequency (RF) sensing framework driven by vision-language models (VLMs). Existing RF sensing methods rely on task-specific designs and provide only partial environmental information, limiting their ability to handle emerging 6G applications. To address this, we propose a generative formulation for RF sensing, where millimeter-wave (mmWave)/terahertz (THz) channel gains are mapped to captions describing multi-level environmental semantics. The framework solves this problem through a complementary design for RF-environment semantic bridging, where a VLM is fine-tuned to leverage its multimodal representations and prompt-conditioned semantic generation capabilities. Hence, different sensing tasks are specified through textual prompts, enabling the framework to handle diverse tasks in a unified manner. For fine-tuning, we introduce prompt-routed low-rank adaptation (LoRA) experts to achieve level-aware adaptation. Simulation results show that, compared with baselines, our framework achieves superior performance with a broader semantic scope, and enables task-unified sensing beyond predefined tasks. Under an unseen sensing requirement, it achieves an average F1-score improvement of 0.17 over the most competitive variant.

Matched filtering is a central result of communication theory, providing the optimal linear receiver when the received waveform satisfies specific assumptions. Practical communication systems rarely satisfy these assumptions, yet learned receivers that outperform the classical matched filter provide little insight into what those improvements reveal about the limitations of the underlying theory. Here we introduce an explainable deformable matched-filter framework in which machine learning is constrained to learn a low-dimensional deformation of the classical matched filter rather than replacing it. Because every learned correction is defined relative to the theoretical matched-filter solution, the deformation becomes a measurable representation of receiver mismatch rather than an unconstrained optimisation. The communication waveform remains processed entirely by the matched filter, while a Kolmogorov-Arnold Network predicts only the deformation from physically interpretable receiver-state descriptors. Using an optical wireless communication testbed spanning ten signalling formats, four impairment classes and 1,600 conditions, we show that learned deformations improve receiver performance, yielding a median relative error-vector-magnitude reduction of 18.1%, while revealing departures from classical matched-filter optimality. Different signalling families occupy distinct deformation regimes, spectral analysis identifies the physical mechanisms underlying receiver mismatch, and latent receiver-state organisation demonstrates that these departures are structured rather than arbitrary.

This paper addresses real-time motion planning for servo systems subject to nonlinear, state-dependent actuator constraints. A reference reshaping method is proposed that combines analytical optimality with high computational efficiency. Using Karush-Kuhn-Tucker conditions, the problem structure is explicitly characterized, and it is shown that the optimal solution lies within a finite set of candidate points. The complete solution set is constructed via closed-form expressions and a small-scale eigenvalue problem, yielding a deterministic algorithm that recovers the exact optimal solution without iterative optimization or root-finding. To address kinematic mismatch induced by aggressive commands, a real-time trajectory compensator is introduced to correct accumulated position error while preserving feasibility. Simulation results demonstrate significant computational speed improvements over existing methods, enabling real-time implementation at frequencies approaching 100 kHz.

The vocal music of each language carries a distinctive sonic identity, even without instrumental accompaniment. We ask whether these differences are measurable and traceable to specific phonemes. To tackle this question, we introduce phoneme-conditional analysis, which isolates the acoustic effect of typologically distinctive phonemes by comparing marker syllables against matched non-marker controls within the same song, holding singer, melody, and genre constant. Across nine typologically diverse languages and thousands of songs, we measure effects along five acoustic dimensions. Song-level profiles built from these effects identify the language of an unaccompanied vocal at 85.5% balanced accuracy in a nine-way classification with folds grouped by artist; whether the separability arises by accumulation of the phoneme-local effects themselves is left open. Our findings suggest that phonological structure leaves systematic and measurable traces in how each language is sung.

In this paper, we investigate the impact of Doppler-induced time scaling in intensity-modulation/direct-detection (IM/DD) multicarrier free-space optical links subject to high mobility. Modeling the Doppler effect as a single-sideband frequency shift applied to the analytic signal waveform, we show that it produces a baseband-equivalent phase rotation, resulting in a common-phase error (CPE), in addition to Dirichlet-kernel inter-carrier interference (ICI), even though the optical carrier phase is removed through photo-detection. The numerical results based on DC-biased optical orthogonal frequency-division multiplexing (DCO-OFDM), show that at small Doppler shifts, the CPE component dominates and can be effectively compensated through pilot-aided phase tracking, while the induced ICI remains rather negligible. At larger Doppler shifts, however, ICI becomes dominant and results in a bit-error-rate floor, highlighting the need for both CPE tracking and ICI mitigation. This is particularly relevant for optical links with small subcarrier spacing and large Doppler shifts, such as those in low-Earth-orbit satellite-to-ground and inter-satellite transmission scenarios.

Obtaining data from neuromorphic sensors and processing it with Spiking Neural Networks is a promising solution to lower the energy cost of artificial intelligence. The current rarity of natively neuromorphic datasets promotes the development of software tools to translate input sensory data into spikes. However, highly bio-mimetic simulators can be challenging to implement on digital hardware. In this work, we evaluate the neuromorphic encoding and subsequent classification of audio into spikes using a non-learnable, high-level, programmable encoder targeting hardware implementation on FPGA. We quantify the pipeline's efficiency with hardware-agnostic metrics based on the quantitative spiking activity. Our study focuses on the simultaneous optimisation of encoder and classifier: the first provides efficient and informative data so that the latter achieves a better performance with an overall lower energy cost at learning and inference. This work introduces the first end-to-end neuromorphic spike-encoding and evaluation of the TIMIT dataset. Our simple feedforward network reaches a classification accuracy of 99.77% on a spike-encoded Heidelberg Digits, overcoming the neuromorphic state of the art on this benchmark dataset.

Large language model (LLM)-based automatic speech recognition (ASR) systems achieve strong performance on conventional speech data by leveraging powerful linguistic priors and multilingual capabilities. However, under challenging conditions, these priors can override acoustic evidence, resulting in unintended translation, instruction execution, repetition, or catastrophic deletion. We propose Likelihood-Constrained Acoustic Reranking (LCAR), a training-free decoding method that improves acoustic grounding while preserving support from the base model. At each decoding step, LCAR first retains tokens whose base-model likelihood falls within a margin of the greedy token, then reranks them using an acoustic compatibility score computed from attention-pooled audio embeddings and the existing LM head. By restricting acoustic intervention to plausible, model-supported alternatives, LCAR requires no additional training, external detector, reference transcript, or auxiliary model at inference. We evaluate LCAR on four LLM-based ASR systems using human-audited TTS and open-source speech challenge suites. At $\delta=0.60$, LCAR removes 38.8–57.1% of detector-identified hallucination failures while largely maintaining WER/CER on standard open-source test sets.

This paper proposes two spectrally weighted STFT loss functions for lightweight streaming speech enhancement, addressing the magnitude over-attenuation in mid-to-high frequency regions caused by the magnitude-phase compensation effect. The proposed sigmoid-weighted loss applies a smooth frequency-dependent modulation to the phase-aware contribution, while the signal-dependent spectrally adaptive loss further conditions the modulation on the ground-truth log-magnitude spectrogram. To evaluate the proposed objectives, we additionally design HyST-Net, a lightweight and competitive backbone with hybrid MHA-GRU spectral-temporal modelling for low-latency streaming scenarios. Experimental results exhibit consistent improvements in high-frequency spectral reconstruction for both losses. The spectrally adaptive loss further enhances the mid-frequency region, resulting in a more balanced spectral reconstruction across the full frequency range.

We derive bounds on manipulating partially coherent microwaves with programmable scattering systems. We first consider the concentration of power from a partially coherent input into a single output port and derive a prototype-aware bound by combining multiport network theory (MNT) with semidefinite relaxation (SDR). We then address general coherency-matrix synthesis, deriving architecture-independent bounds on fidelity and useful strength, prototype-specific scalar refinements thereof, and fully prototype-aware SDR bounds on the useful-strength–fidelity Pareto frontier. The prototype-aware formulations account for mutual coupling, loss, discrete tunability, and static scattering. We evaluate the bounds on four experimental RIS-parametrized MIMO systems with up to 100 1-bit-programmable elements, using proxy-MNT models estimated from measurements. The resulting bounds are generally tight compared with feasible discrete-optimization outcomes. Interestingly, for coherency synthesis, the simpler bounds can in some cases be tighter than the fully prototype-aware bounds, highlighting their complementarity. Our results provide certified limits for wave-domain processing and harvesting of partially coherent microwaves.

This paper develops an optically movable Rydberg atomic quantum receiver (RAQR), in which the probe and coupling beams are steered within each vapor cell to dynamically reconfigure the effective radio-frequency (RF) sensing position without mechanical actuation. A closed-form equivalent baseband model is derived by separating the atomic transduction coefficient, optical steering phase, and cell-center array response into distinct factors and the accuracy of the resulting model is validated against numerical solutions of the Lindblad master equation. Based on the derived model, we reveal two complementary channel-shaping mechanisms, including intrinsic beam-pattern shaping through RF-to-optical transduction and per-cell phase control enabled by optical displacement. To further exploit these capabilities, a non-convex sum-rate maximization problem is formulated over the optical positions and local oscillator design and solved via an alternating optimization framework with analytical gradients. Simulation results validate the derived model and demonstrate substantial performance gains enabled by optical movability, highlighting its potential as a programmable receiver architecture for future wireless networks.

In doubly selective channels, the joint delay and Doppler dispersion generally induces coupling among transmitted symbols, thereby increasing receiver equalization complexity. Nevertheless, by using appropriately designed waveforms, channels with certain delay-Doppler (DD) supports can be diagonalized for one-tap equalization. The whole picture of such DD supports and their corresponding waveforms is still unclear, except for several examples identified in literature. In this paper, under cyclic-prefix (CP)-based block transmission and assuming that the modulation waveforms form an orthonormal basis, we identify all such channel supports by an elementary expression, and derive the corresponding waveforms in closed form.

Wireless-enabled physical artificial intelligence (physical AI) systems call for a shift from reliable data delivery to reliable inference of physical events. The relevant question is not only whether packets arrive, but whether the set of cues available at the decision node, i.e., the evidence, is sufficiently timely and informative to support reliable inference about the event. Accordingly, this paper develops a framework in which event-inference reliability (EIR) is determined jointly by cue informativeness, cue availability, and temporal admissibility. The latter is determined by the downstream task requirement and represented through the usefulness horizon. We define event-inference error ratio (EIER) as the normalised residual event uncertainty after incorporating admitted cues, and EIR as the corresponding normalised uncertainty reduction, both conditioned on decision-node context. We further distinguish the evidence-limited Bayes benchmark from operational performance of a particular inference engine and derive an entropy-based lower bound on the minimum achievable event error from the same decision-node information. The framework then enables an event-aware wireless design interface for cue prioritisation, cue-reliability allocation, and event-inference coverage characterisation. A multiclass indoor activity-inference study combining empirical cue likelihoods with wireless delivery instantiates the framework and demonstrates how it characterises EIR under finite usefulness horizons.

Microwave linear analog computers (MiLACs) have emerged as a promising architecture for energy-efficient wireless communications by shifting signal processing from the digital to the analog domain using tunable impedance networks. Several MiLAC architectures have recently been proposed, including the single-layer MiLAC (SLM), two-layer MiLAC (TLM), and hybrid digital-MiLAC (HDM). In this paper, we investigate the energy efficiency (EE) of these architectures. Specifically, we formulate EE maximization problems for the SLM, TLM, and HDM architectures under transmit power, user rate, and architecture-specific constraints, and develop a dimensionality reduction technique with successive convex approximation (SCA)-based algorithms to efficiently solve the resulting non-convex problems. We further derive a computationally efficient solution for EE maximization based on a search over only $(K+1)$ closed-form candidate solutions, where $K$ is the number of users, and analyze the asymptotic EE in the large-antenna regime under negligible quantization noise. Our analysis shows that the EE of the SLM, TLM, and HDM architectures scales as $\ln(N)/N^2$, whereas conventional digital beamforming (DBF) scales as $\ln(N)/N$, where $N$ denotes the number of transmit antennas. Despite these different scaling laws, the faster EE decay of MiLAC-based architectures becomes relevant only at very large antenna dimensions, typically involving thousands of antennas, while they maintain superior EE over practically relevant finite antenna regimes. Finally, we derive an approximation of the number of antennas required to achieve the maximum EE for each architecture and the corresponding EE. Simulation results show that, over practically relevant antenna regimes, MiLAC-based architectures achieve substantially higher EE than conventional fully digital and hybrid analog-digital beamforming.

Fixed linear compression can preserve local Fisher information for a sinusoid yet destroy global frequency identification, while a dark response can make local estimation nonregular. We study both failures for a single complex tone observed through a fixed complex-linear sketch with unknown nonzero complex gain and pre-sketch white Gaussian noise. Near an isolated analytic dark frequency, we separate radial signal vanishing from optimized projective contact between the two signed frequency branches. When the gain magnitude is constrained to a fixed nondegenerate interval, finite contact is equivalent to local quotient identifiability and minimax consistency. The sharp mean-square-error rate is determined by the sum of the radial and contact orders; infinite contact produces exact local aliases. Globally, we formulate a worst-frequency efficientinformation objective on whitened row spaces. We solve it exactly for every even output rank, prove uniqueness and quantitative rigidity of the symmetric-edge projector, and solve the evenaperture co-rank-one case. The rank-two optimum is aliased, and a winding obstruction gives a positive lower bound on the information price of global identification. From three outputs onward, sketches that are globally identifying with an immersive projective response are open and dense and have zero price at the level of suprema; for every even rank of at least four, the exact optimizer has this property. Thus local information preservation, singular recoverability, and global identifiability obey distinct compression laws within one estimation model.

Deepfake voice detection suffers from poor generalization across unseen domains. While Audio Large Language Models (ALLMs) show promise, the modality gap between continuous audio embeddings which capture the subtle acoustic details necessary for deepfake detection and the semantic space of LLMs remains a critical, underexplored bottleneck. We address this by benchmarking diverse audio encoders integrated with Qwen LLMs (0.5B to 7B parameters). First, we demonstrate that fine-tuning the LLM alone risks out-of-domain overfitting, making a frozen LLM a stronger, resource-efficient baseline. Second, to explicitly bridge the modality gap, we introduce a cross-modal prompting strategy that injects linguistic-knowledge-driven acoustic features (via openSMILE) as structured text tokens. This explicit textual grounding not only enhances the frozen baseline but also makes LLM fine-tuning more effective. Ultimately, our approach demonstrates state-of-the-art resilience on the out-of-domain ITW and MLAAD benchmarks, yielding over 16.2% absolute improvement in Macro-F1 over existing ALLM baselines while maintaining competitive in-domain performance. All models reported in this work are \href{https://huggingface.co/01Yassine/AudioLLM-Deepfake-Detection}{publicly available}.

The increasing integration of Internet-of-Things (IoT) technologies for monitoring and control of inverter-based resources (IBRs) has expanded the attack surface, enabling stealthy manipulation of controller parameters through vulnerable supervisory control interfaces. Phase-locked loops (PLLs) emerge as prime targets, as they interact with all control loops and critically influence the dynamic response of grid-following (GFL) converters. This paper analyzes the underlying threat model to elucidate the mechanisms enabling such stealthy behavior and conducts a thorough stability and transient response analysis to characterize how PLL tampering can degrade system performance without necessarily destabilizing the system. The results reveal critical interactions among frequency estimation, control, and synchronization that can significantly reduce stability margins. To counter the stealthy nature of these attacks, a modified PLL is proposed that exposes gain variations through shifts in its equilibrium points while preserving conventional PLL performance. Experimental results validate the effectiveness of the proposed approach in detecting PLL cyber-attacks.

Accurate wireless localization is a key enabler for 6G networks, yet remains challenging under diverse and rapidly changing propagation conditions. Model-based methods degrade when multipath channels are non-resolvable and model mismatches occur, while supervised deep learning demands large labeled datasets and generalizes poorly to new deployments. Inspired by foundation models (FMs) in language and vision, this article presents a unified framework for FM-based wireless localization that learns transferable channel representations from large-scale unlabeled channel state information and adapts to new environments with minimal or even no supervision. We review the fundamentals of FMs, compare the FM paradigm with existing localization approaches, and introduce a three-stage framework spanning large-scale pretraining, localization-oriented fine-tuning, and context-augmented inference, together with the location-aware applications it enables. Ray-tracing-based case studies show improved positioning accuracy and cross-environment generalization. Finally, we present an outlook on key research directions toward AI-native networks for wireless localization.

Beamforming methods in millimeter-wave (mmWave) cell-free massive multiple-input multiple-output (CFmMIMO) systems require accurate channel state information (CSI), whose acquisition entails significant training overhead. This paper shows that fully digital cell-free mmWave beamforming can be effectively learned from sub-6 GHz CSI using a graph neural network (GNN). Specifically, we represent a CFmMIMO system as a wireless graph, and the GNN is trained to approximate beamformers that maximize the downlink sum-rate based on the available sub-6 GHz CSI. A message-passing mechanism is proposed to capture inter-user interference and inter-base-station cooperation across different network topologies. Simulation results demonstrate that the proposed sub-6 GHz-assisted GNN-based beamformer achieves competitive and often superior sum-rate performance compared to classical baselines that rely on full mmWave CSI.

Flexible sensing arrays are commonly used in underwater acoustic networks, but suppressed by unpredictable geometric deformations. Existing array shape self-calibration methods often estimate individual element positions separately, leading to a high dimensional optimization problem over long arrays. To address this problem, this paper proposes a Bayesian Optimization-assisted Geometry Estimation (BOGE) strategy operating with a hierarchical optimization process and a physics-informed parametric model for array geometry correction. BOGE formulates array shape self-calibration as an optimization problem, where candidate geometries are evaluated by the noise subspace residual. We perform Bayesian optimization to configure the physics-informed parametric model and then refine the selected geometry through numerical optimization. Empirical results show that BOGE achieves lower mean geometric root mean square error (RMSE) than the benchmark methods across a wide range of noise levels. On the public SWellEx-96 dataset, BOGE achieves a geometric RMSE of $0.659$ meters at $166$ Hz. A lake trial further shows that BOGE provides fixed source localization and moving target tracking performance comparable to the comparison methods.

By focusing on algorithmic stability as a means of establishing out-of-sample bounds, we provide a system-theoretic interpretation of generalization in learning-enabled dynamical systems arising in data-driven optimization and feedback control approximation. Given two neighboring datasets, we specifically model sample replacement as an exogenous disturbance acting on a sensitivity system, while the incremental behavior of the data-dependent operator is encoded through an integral quadratic constraint. By relying on dissipativity arguments, we establish a matrix inequality-based certificate and a uniform stability bound that separates the one-sample sensitivity of the learned operator, and an algorithm-dependent dynamical gain. The latter can then be optimized, offering a tractable tool for certifying and comparing generalization capabilities of learning dynamics. We show that our results recover classical ones for gradient descent, apply naturally to momentum-based methods such as heavy-ball and Nesterov acceleration, and extend to data-driven control.

Resting-state functional magnetic resonance imaging (rs-fMRI) functional connectivity (FC) matrices are widely used for individual-level prediction, but strong performance within one cohort may not generalize to a new cohort. We ask whether within-dataset performance remains when the test data come from an entirely held-out rs-fMRI dataset. Each scan is represented as a regularized symmetric positive definite (SPD) correlation connectome, which allows methods to use the geometry of the SPD manifold. We introduce a reproducible age-prediction benchmark across six rs-fMRI datasets: COBRE, ADNIDOD, Cam-CAN, ABIDE, OASIS-3, and ADNI. The benchmark compares a vectorized correlation baseline, Tangent-Space Ridge, SPDNet, and split-wise Riemannian harmonization under within-dataset GroupKFold, pooled GroupKFold, and leave-one-dataset-out (LODO) evaluation. Within-dataset and pooled GroupKFold results are substantially more favorable than LODO results. When an entire dataset is held out, prediction error increases, differences among methods narrow, and performance is strongly affected by age-range mismatch and cohort heterogeneity. The benchmark provides common inputs, model settings, data splits, and analysis scripts so that future SPD matrix learning methods can be evaluated under the same external-validation protocol.