The operational reliability of a high performance marine vessel depends critically on the health of its marine propulsion systems, which are increasingly subjected to diverse operational loads and environmental stressors. This paper proposes a robust mathematical framework for non-linear state-space forecasting of marine engine parameters using adaptive-window multi-particle stochastic differential equations. Traditional time-series models such as Vector Autoregressive Integrated Moving Average, often fail to capture the inherent stochasticity and transient dynamics of complex systems due to their reliance on fixed-window linear assumptions. To address this, we develop a dual-layered estimation approach: first, an adaptive lookback mechanism dynamically adjusts the learning window size based on the instantaneous drift magnitude, ensuring responsiveness during non-stationary regimes. Second, a Multi-Particle ensemble is evolved via Euler-Maruyama discretization, where each particle trajectory represents a stochastic realization of the system state. To refine the ensemble mean and mitigate the "noise-chasing" behavior of raw estimators, a Girsanov transform induced change of probability measure is implemented, assigning higher probabilistic weights to particles that align with the physical drift. Theoretical evaluation and empirical benchmarking demonstrate that the proposed adaptive SDE framework significantly outperforms classical statistical baselines in multi-step prediction stability and computational efficiency. The model provides a scalable, "grey-box" solution for real-time risk quantification in systems characterized by high-frequency volatility and non-linear transitions.

Future missions to Mars and Venus will make use of aerobraking and aerocapture in order to gain mass through the saving of fuel at planetary arrival. So far only aerobraking has been demonstrated, if the Mars Premier project has paved the way for aerocapture, no demonstration was performed due to the project interruption. The use of these techniques induces additional constraints for planetary probes, since additional heating and mechanical loads have to be carefully managed. Moreover, aerocapture requires a high level of accuracy for the Guidance Navigation and Control aspects, since a pass at an altitude of the atmosphere with a different density could lead to the vehicle destruction. This document surveys the existing state-of-the-art on inflatable devices (including ballutes, sails, or inflatable heat-shield capsule) for orbital manoeuvres at planetary arrival.

Future wireless systems are expected to employ extremely large-scale multiple-input multiple-output (XL-MIMO) arrays at high carrier frequencies, where near-field propagation makes the channel depend jointly on angle and distance. The resulting short coherence intervals make channel state information acquisition challenging, motivating blind channel estimation and data detection (B-CE-DD). In this paper, we propose a two-stage B-CE-DD framework for uplink near-field XL-MIMO systems. First, we formulate the problem as the recovery of user-specific rank-one channel-data products from a superimposed received signal using a polar-domain sparse channel model and a low-dimensional data subspace model. Building on this formulation, we develop an on-grid blind orthogonal matching pursuit (B-OMP) algorithm that exploits polar-domain sparsity to iteratively identify the dominant angle-distance components and estimate the corresponding channel-data products, followed by an off-grid refinement stage based on block-coordinate descent (BCD) that optimizes the angle and distance parameters in the continuous polar domain. Numerical results show that the proposed B-CE-DD framework combining B-OMP and BCD significantly improves the symbol error rate compared with a pilot-based baseline employing zero-forcing beamforming, particularly at low signal-to-noise ratio and when the number of data symbols is small relative to the length of the coherence interval.

This paper proposes a systematic framework to assess the small-signal stability of power systems with high shares of grid-following inverter-based resources (IBRs) under varying controller parameters and operating conditions. Stability manifolds are introduced to identify controller-parameter regions that ensure stability across multiple scenarios. Full-network linearization and eigenvalue analysis are combined with adaptive sampling based on probabilistic support vector machine classification to approximate stability boundaries efficiently, while surrogate optimization identifies feasible initial controller settings meeting bandwidth and phase-margin constraints. The approach is validated on a modified Cigré European HV network benchmark with 50 operating scenarios and increasing inverter penetration. Results show that stability sensitivity grows with inverter share, interactions among IBRs reshape admissible parameter regions, and simplified equivalent-network models may overlook critical system-level limitations. The framework supports stability-oriented controller design and interconnection studies in converter-dominated systems.

Early prediction of respiratory failure is critical for timely clinical intervention in intensive care units. Existing electronic health record (EHR)-based models can continuously monitor physiologic deterioration, but they may not fully capture pulmonary pathophysiology reflected in chest radiographs (CXRs). In this study, we ask whether CXR information improves prospective prediction of invasive mechanical ventilation beyond EHR signals alone. We develop a gated multimodal framework that integrates structured EHR time-series data with CXR foundation-model representations. The gating module adaptively controls the contribution of imaging features based on patient-specific clinical context, allowing the model to selectively rely on imaging information when it is informative. We prospectively evaluate the framework for predicting invasive mechanical ventilation within 24 hours in ICU patients and compare it with an established EHR-only model (this http URL), physician predictions obtained at matched clinical time points, and alternative multimodal variants. The gated multimodal models achieved higher discrimination than the EHR-only baseline, with AUROC values of 0.860 and 0.858 using REMEDIS and MedInsight CXR representations, respectively, compared with 0.752 for this http URL. Relative to physician predictions, the multimodal framework substantially improved sensitivity while maintaining favorable specificity. Compared with the EHR-only model, multimodal integration increased specificity and positive predictive value, suggesting that CXR information can refine risk estimation in selected patients. These findings support adaptive multimodal fusion as a practical strategy for incorporating imaging into prospective respiratory failure prediction.

In recent years, increasing aerospace safety requirements have intensified the demand for reliable structural damage detection. This work presents an Operational Modal Analysis approach for accurate modal parameter estimation, with an application to space structure monitoring. The proposed System Identification (SI) method innovatively combines the Natural Excitation Technique (NExT) with the Fast and Relaxed Vector Fitting (FRVF) algorithm, which uses an iterative least-squares optimisation. A preliminary validation is first carried out on a numerical beam model, comparing results with analytical solutions and the established Natural Excitation Technique with Eigensystem Realisation Algorithm (NExT-ERA) and Stochastic Subspace Identification with Canonical Variate Analysis (SSI) methods. Then, operational validation is performed on real acceleration data from the Space Acceleration Measurement Systems aboard the International Space Station. Identified vibration modes from NExT-FRVF and NExT-ERA show comparable results after signal processing, with mode consistency assessed by repeated occurrence and physical interpretation, while SSI fails to identify most. The output-only algorithm proves to be highly reliable, outperforming benchmark methods under noisy conditions on a numerical system and offering reliable identifications on the experimental data.

In cell-free massive multiple-input multiple-output systems, downlink power control is essential to ensure uniformly high service quality across users. Existing methods range from centralized iterative approaches requiring global channel knowledge and supervised training, to simpler distributed strategies such as fractional power control that rely on local information but perform poorly in terms of fairness. This letter proposes an unsupervised, physics-informed framework that directly optimizes max-min fairness without requiring optimal labels or user position information. The method is inherently scalable in the number of user equipment, does not require retraining when the user population changes, and can be extended to achieve full scalability with respect to both access points and users. Numerical results show that it nearly doubles the worst-user spectral efficiency compared to existing scalable schemes.

Online continual learning (OCL) enables real-time adaptation to new data, making it crucial for dynamic robotic applications. However, its practical deployment is hindered by memory constraints in resource-limited systems, which affect key trade-offs in training latency, plasticity, and stability. Unlike offline parameter tuning, which cannot account for the dynamic shift in memory pressure and workload complexity as OCL progresses, an online and self-adaptive approach is essential for robust on-device deployment. This paper proposes Orion, a holistic framework designed to co-optimize training latency, plasticity, and stability of state-of-the-art OCL models under strict memory constraints, enabling feasible on-device deployment. At its core, Orion leverages URGE, a unified runtime indicator grounded in the ``Buckets effect'' principle that system performance is bounded by its scarcest resource, to dynamically reallocate memory across OCL components by jointly coordinating batch processing, replay buffers, and optimization strategies at both the OS and application level. Furthermore, Orion introduces system-level data prefetching techniques to maximize efficiency. A system prototype of Orion has been implemented using the widely adopted \texttt{Avalanche-lib} and thoroughly evaluated across a diverse range of OCL algorithms, benchmarks, and hardware platforms commonly used in autonomous robotic applications. To further demonstrate its practical utility, Orion is integrated into a realistic autonomous navigational robot powered by OCL. The results show that Orion achieves significant training speedups while maintaining balanced performance and effectively adapting to various scenarios, all with minimal runtime, memory, and energy overhead, making Orion a practical solution for on-device continual learning.

The pinching-antenna systems (PASS), which dynamically activate and relocate the pinching-antennas (PAs) along the dielectric waveguide, offer unprecedented potential for integrated positioning and communication. The multi-waveguide-based uplink positioning approaches for indoor environments are first proposed in this paper, and the downlink communication performance is analyzed. Two possible scenarios, multi-waveguide single-PA (MWSP) and multi-waveguide multi-PA (MWMP), are considered under the assumptions of line-of-sight channels and a single, stationary user. For the MWSP scenario, the received signal strength indication (RSSI)-based ranging method and the MWSP-based least square (LS) positioning algorithm are developed. To gain deeper insights, a comprehensive error analysis of the LS positioning algorithm is conducted. Subsequently, for the MWMP scenario, the closed-form expression of the superposed signal is derived. According to the signal power, the MWMP-based grid search algorithm is proposed and the estimation error of proposed algorithm is analyzed. Then, based on the user's positioning result, the PAs are relocated to provide downlink communication service, and the achievable data rate of MWSP and MWMP scenarios are analyzed. Numerical results validate the correctness of our analysis, which show that: i) For the MWSP scenario, a smaller geometric dilution of precision (GDoP) leads to a lower average positioning error. Furthermore, even when the GDoP is large, the regions where the distances to PAs are nearly equal achieve the best accuracy. ii) For the MWMP scenario, non-parallel waveguide deployment improves positioning accuracy, although errors increase with the number of PAs. iii) The noise has a serious double-impact on data rate. There is a trade-off between positioning accuracy and communication performance.

With the rapid development of satellite communication and navigation, there is an urgent need to integrate both technologies to achieve reliable communication and precise navigation services within the same satellite system. By combining multi-/uni-cast (MUC) and non-orthogonal multiple access (NOMA) technologies, we propose a novel MUC-NOMA-based integrated navigation and communication (INAC) signal structure, in which the navigation and communication signals share a common pseudo noise (PN) sequence, thereby integrating satellite communication and navigation at the signal level. According to different power allocation strategies, two scenarios are defined: multi-cast-oriented (MO-) INAC and uni-cast-oriented (UO-) INAC, where a greater portion of power is assigned to either the multi-cast or the uni-cast signal, respectively. To mitigate co-channel interference, we employ successive interference cancellation (SIC) at the receiver and design a signal processing algorithm for the proposed INAC signal. Then, closed-form expressions are subsequently derived for the bit error rates (BER) of both the navigation and communication signals, along with the positioning accuracy of the navigation signal. To gain further insights, the impacts of power allocation factors and communication rates are evaluated. Our analysis results show that: i) In the MO-INAC scenario, the positioning and BER performance of navigation signal are excellent when more power is assigned to the multi-cast signal; ii) In the UO-INAC scenario, interference in the shared resources is reduced when more power is assigned to the uni-cast signal; iii) The ranging accuracy decreases as the communication data rate increases. Numerical results confirm the superior BER and positioning accuracy of the MO-INAC scenario for MEO satellites.

Safety filters constructed from control barrier functions (CBFs) are commonly appended to pre-trained neural network controllers to enforce safety requirements. However, this decoupled design with hand-tuned, fixed CBF parameters often fails to adapt to the underlying controller, yielding overly conservative solutions. Thus, given a valid CBF, we address these limitations by jointly learning a neural network controller and neural-network-parameterized CBF parameters, enforcing the resulting affine safety constraints by construction and avoiding an online quadratic program (QP) safety filter at run time. To further improve computational efficiency and scalability, we introduce a lightweight projection architecture that enforces constraints without full constraint enumeration. Extensive simulation evaluations demonstrate reliable, scalable safety constraint satisfaction at reduced computational cost.

In Global Navigation Satellite System (GNSS)-denied environments, terrestrial signals of opportunity (SoOP) offer an alternative for positioning, but synchronization impairments such as clock offsets, drift, and multipath limit performance. This paper proposes a receiver-centric multi-channel time-difference-of-arrival (TDOA) localization framework based on Digital Audio Broadcasting (DAB) signals. The method exploits the DAB null symbol for coarse timing and the phase reference symbol (PRS) for fine synchronization, followed by sub-sample time-of-arrival (TOA) estimation. A double-difference formulation removes inter-receiver clock offsets, while a peak-to-sidelobe ratio (PSR)-based weighting improves robustness. A bias correction step mitigates errors due to multipath. Finally, a coordinated-turn extended Kalman filter (CT-EKF) further refines position estimates. Results show improved accuracy over conventional TDOA with Gauss-Newton estimation, especially in challenging conditions.

Learning-based methods for synthesizing controllers have gained popularity due to their high expressiveness and strong empirical performance. However, in safety-critical scenarios such as autonomous driving, robotics, and power systems, empirical performance alone is insufficient, and formal verification of controller properties such as stability and safety is highly desirable. Unfortunately, many prior verification approaches are either tied to specific structural assumptions on the system or the certificate, making them difficult to transfer across settings, or suffer from poor scalability on higher-dimensional neural network systems. In this tutorial, we present a unified framework that aims to mitigate this gap via bridging control with the state-of-the-art neural network verifier $\alpha,\!\beta$-CROWN (alpha-beta-CROWN). At its core, $\alpha,\!\beta$-CROWN is a general-purpose bounding engine for nonlinear functions represented as computation graphs: given an input domain, it can produce certified bounds and explicit linear relaxation of the nonlinear function. These certified bounds are useful on their own for tasks such as reachability analysis, and they also provide the foundation for more complex routines that perform satisfiability checking and optimization. More specifically, many control problems reduce to verifying real-valued inequalities over a state domain (e.g., Lyapunov theory). Consequently, $\alpha,\!\beta$-CROWN enables scalable verification of such conditions by computing tight bounds and recursively partitioning and pruning subdomains based on the bounds. Thanks to GPU parallelization, this pipeline demonstrates superior scalability on verification and optimization problems that are challenging for traditional approaches. In this tutorial, we discuss the basics of $\alpha,\!\beta$-CROWN and introduce its application to various control-related tasks.

The large-scale Electric Vehicle (EV) integration into the electricity grid has initiated significant challenges to grid stability issues due to dynamic loadability events. Although electric vehicle impacts on distribution systems are well studied, transmission-level investigations remain limited. In this research paper, case scenarios of EV load models as charging stations have been considered for stability analysis (Voltage and Frequency Stability) to address EV operation on the transmission grid. It is also noted that the operation of EV stations due to their high loadability causes more stability complexities to the grid compared to other loads in a power network. Simulations have been conducted on two different power networks of the IEEE-9 and IEEE-39 bus test systems, respectively.

Stochastic model-predictive control (SMPC) has evolved to a powerful framework for the control of stochastic dynamical systems. SMPC utilizes a probabilistic uncertainty description to provide a systematic trade-off between the control objective and constraint satisfaction in a statistical sense. However, the majority of existing SMPC methods face challenges related to computational tractability due to the need for stochastic inference. Approaches that apply accurate inference are computationally demanding, which can lead to serious limitations in the implementability of these methods. Hence, in practice, the uncertainty propagation and the resulting distributions are typically approximated, e.g., by Gaussian distributions. These approximations promote computational efficiency, but are often too conservative, becoming a limiting factor in the representation of stochastic state evolution and the implied guarantees. To overcome this fundamental limitation of SMPC approaches, we propose a novel formulation based on the meta-state-space (MSS) representation of stochastic dynamical systems. The proposed MSS-based SMPC scheme offers a computationally efficient way to forward propagate the uncertainty with a flexible and highly accurate approximation of the probabilistic system description. With the presented method, the entire output probability density function can be directly shaped, which is unprecedented among existing SMPC techniques. Finally, we provide a detailed theoretical analysis and demonstrate the effectiveness of the proposed methodology via an extensive simulation study.

Deploying reinforcement learning in safety critical domains, from autonomous vehicles to medical decision support, is constrained by failures arising when systems encounter unfamiliar conditions. We argue that the fundamental bottleneck is not individual challenges like changing dynamics or incomplete observations, but their synergistic interaction, which we term the Epistemic Trap: agents cannot estimate their state without knowing system dynamics, nor learn dynamics without accurate state information. Proof-of-concept experiments in simulated locomotion reveal that combining these uncertainties causes failures far worse than either challenge alone, a 77% performance degradation against the 46% by adding the individual effects, demonstrating compounding failure modes that conventional methods overlook. Such approaches adopt a passive epistemic stance that cannot resolve this coupled uncertainty. We propose reframing safety as an information problem, introducing an Adaptive Safety Architecture built around three contributions: the Compound Uncertainty Coefficient ($\kappa$), a mutual information based metric that quantifies state dynamics coupling and is computable online without full joint belief inference; information seeking policies governed by a MaxInfoRL objective that actively probe system dynamics; and regime-adaptive safety constraints that tighten as epistemic coupling rises. This paradigm shift, from passive robustness to active perception, offers a principled path toward decision making systems that operate under uncertainty, recognize their own ignorance, and act strategically to resolve it.

We study the sample complexity of policy gradient for log-growth control -- the problem of learning, from observed state transitions, a feedback gain that optimally stabilizes a scalar linear system driven through a multiplicative-noise actuation channel. The objective $J(K) = \mathbb{E}[\log|1+BK|]$ is the top Lyapunov exponent of the closed loop. This problem carries a structural difficulty we call the cusp obstruction: the optimal gain $K^*$ always places the noise singularity $b_{\rm sing}(K) = -1/K$ in the interior of the support. At this singular optimum the policy gradient exists only as a Cauchy principal value, not as a Lebesgue integral, and the natural single-sample gradient estimator has infinite variance. Standard first-order stochastic-optimization analysis is thus inapplicable at the optimum, and merely smoothing the objective does not resolve the difficulty. The obstruction, however, has an exploitable symmetry: the Cauchy kernel is an odd function of the displacement from the moving pole, so pairing each observation with its reflection through the pole cancels the divergent part. This one cancellation simultaneously controls the population curvature, the gradient-estimator variance, and the bias incurred when the noise density is estimated. Combining these bounds with a closed-form single-transition gradient oracle, we prove that projected mini-batch policy gradient, initialized in any compact subset of the stabilizing region, attains total sample complexity $\tilde{O}(1/\eta)$ when the noise density is known and $\tilde{O}(\eta^{-(2s+1)/(2s)})$ when it must be estimated, for $C^s$ noise densities with $s \geq 2$.

Neural cellular automata (NCA) provide a lightweight alternative to encoder-decoder segmentation networks. However, it can be difficult to decide when a prediction should be trusted. Here, we study uncertainty estimation for NCA-based medical image segmentation without modifying the underlying architecture or retraining the model. Our approach is motivated by viewing the NCA as a dynamical system where convergent attractors correspond to confident predictions. Concretely, we propose resilience, a simple measure that leverages the intrinsic iterative structure of NCAs by probing the stability of the final prediction under small perturbations of the automaton state. Predictions that return to the same solution are deemed confident, while those that change substantially are flagged as uncertain. We evaluate uncertainty by its ability to predict segmentation quality using selective prediction metrics ($\Delta$Dice@90 and AURC) and ranking metrics (AUROC and AUPRC). Across multiple medical segmentation benchmarks, resilience identifies failure cases more reliably than baselines, improving trust and safety in NCA-based models.

Physical layer security in reconfigurable intelligent surface (RIS)-assisted wireless systems can be improved through coordinated control of signal transmission and RIS configuration. In this work, the base station simultaneously transmits the communication signal (CS) and artificial noise (AN) in the presence of a potential eavesdropper. The RIS is partitioned into two groups of reflecting elements, where a portion enhances the desired CS toward the legitimate receiver, while the remaining elements contribute to AN transmission. Two key parameters govern the system design: a transmit power allocation factor between CS and AN, and an RIS element allocation ratio controlling the partitioning of the reflecting elements. An iterative binary phase optimization strategy is employed to enhance the received signal power at Bob while degrading Eve's reception. Simulation and experimental results demonstrate that proper joint design significantly improves the achievable secrecy capacity.

Ultrasound Localization Microscopy (ULM) has presented great potential in functional imaging, benefiting from its ability to reconstruct deep microvasculature. However, the hemodynamic reconstruction is compromised by sparsity in the ULM data, as a limited number of MB tracks cannot sample the complete speed profile in one vessel. Here, we propose to reconstruct hemodynamics using sparse ULM velocity maps by solving a laminar flow model through stochastic variational inference. In addition to vascular geometry and flow velocity maps, the proposed method generates two new ULM maps - a pressure gradient map and a map describing uncertainty of the estimation. By investigating the effect of sparsity in ULM maps on the quantification and visualization of hemodynamics, we demonstrate the effectiveness of the proposed method in dealing with sparse ULM maps via simulations and 3D rat brain imaging. Accurately reconstructing a broad range of hemodynamic parameters and associate uncertanties using sparse ULM data may help detect subtle and dynamic brain activity.

Multimodal signals on sensor networks are commonly modeled under the twofold graph assumption (TGA), which represents spatial structure and inter-modality relations as two separate graphs. Existing TGA-based signal restoration methods, however, either assume the graphs are known or restrict edge weights to be non-negative, preventing them from capturing negative inter-modal correlations. We address both limitations by formulating joint signal restoration and twofold graph learning as MAP estimation under a matrix normal prior, where the spatial and modality graph Laplacians appear directly as precision matrices. The resulting non-convex objective is solved by alternating minimization: The signal is updated via conjugate gradient applied to the arising Sylvester-type linear system; the graphs are updated via primal-dual hybrid gradient (PDHG). We further propose a method to estimate the signed structure of the modality graph from the dominant eigenspace of a complementary kernel matrix, which is then used in PDHG to update edge magnitudes. These iterative solvers are then unrolled into a feedforward network, with regularization weights and step sizes as layer-wise trainable parameters. Experiments on synthetic multimodal graph signals and a real Japan meteorological dataset confirm that the proposed method outperforms existing baselines across a range of noise levels and missing-data patterns.

In this work, we introduce a real-time capable algorithm for considering monotonicity assumptions for recursive Gaussian Process regression (RGP). Therefore, we present how to efficiently calculate the RGP gradients online. Then, we utilize an extended Kalman filter and pseudo-measurements in combination with a ReLU pseudo-measurement function to enforce soft inequality constraints. This work builds upon a previously published conference paper with the same goal and a similar fundamental approach. Opposite to our previous work, however, we now use an exact covariance calculation for the RGP gradients. Furthermore, we also present a real-time optimized version of this algorithm with less simplifications compared to the previously published version. These and several other algorithmic innovations lead to an algorithm with greatly improved numerical robustness. The algorithm is validated and compared to its previously published version for a 2D numerical example. The paper is concluded with a successful experimental validation of the developed algorithm for the monotonicity-preserving learning of pneumatic valve characteristics for the control of a pneumatic system, leveraging a partial input - output linearization.

Incentive-based load curtailment unlocks critical demand-side flexibility but is hindered by the limited knowledge of private user parameters and the inherent nonsmoothness of responses due to physical device constraints. We address this via a constrained bilevel optimization framework and propose the Bi-ZOL (Bilevel Zeroth-Order Learning) algorithm. Unlike conventional black-box methods, Bi-ZOL exploits the bilevel structure to decompose the hypergradient, integrating the exact analytical information of the SO's objective with a zeroth-order estimate of the unknown response sensitivity. This structural decomposition-based learning method mathematically smoothes the nonsmooth response landscape and reduces hypergradient estimation error. We provide theoretical convergence guarantees to an approximate stationary point and demonstrate through simulations that Bi-ZOL achieves near-optimal performance.

High-quality speech coding at low bitrates is crucial for bandwidth-constrained applications, yet remains challenging due to the severe loss of quality-critical information in highly compressed representations. To overcome this challenge, we propose CFMDCTCodec, a low-bitrate neural speech codec that operates entirely in the modified discrete cosine transform (MDCT) domain. CFMDCTCodec integrates a lightweight encoder-quantizer-decoder-style MDCT-spectral codec with a noise-prior-aware, conditional-flow-matching (CFM)-based MDCT-spectral enhancer. Within this framework, the codec serves as a base module that compactly discretizes the MDCT spectrum extracted from speech and produces an initial coarse reconstruction, while the enhancer further restores fine-grained spectral details. The enhancer improves the decoded MDCT spectrum by integrating a conditional MDCT velocity-field filter with an ordinary differential equation (ODE) solver, under the guidance of an MDCT-derived magnitude-adaptive noise prior, aiming to emphasize perceptually significant high-energy regions while stabilizing low-energy and silent regions. Finally, the enhanced MDCT spectrum is reconstructed into the decoded speech using the inverse MDCT. When optimizing CFMDCTCodec, we adopt a unified non-adversarial training strategy that jointly combines reconstruction, quantization and CFM objectives. Both objective and subjective evaluations show that CFMDCTCodec outperforms competitive baselines in low-bitrate regimes, e.g., 0.65 kbps, while approaching the perceptual quality of large-scale codecs with significantly fewer parameters and computations.

Wi-Fi sensing has emerged as a versatile tool for tasks such as localization, gesture recognition, and vital-sign monitoring, enabling applications from smart environments to personalized healthcare. However, sensing accuracy often significantly degrades when pretrained models are deployed across different commodity receivers. We present the first systematic comparison of Channel State Information (CSI) across diverse Commercial Off-The-Shelf Wi-Fi sensing platforms. Using a unified experimental setup delivering precisely precoded signals simultaneously to multiple receivers, we isolate receiver-specific variability. We find that dominant cross-device differences arise from Automatic Gain Control and consistent subcarrier nonlinearities. We propose a simple gain-alignment preprocessing step, recovering most of the lost accuracy (up to 75%) in cross-device Human Activity Recognition model deployments. Without preprocessing, model accuracy sharply drops-effectively breaking practical deployments. Additional analyses reveal measurable inherent differences in receiver faithfulness, sensitivity and noise. While these receiver-induced differences do not significantly affect robust sensing tasks such as Human Activity Recognition, they become relevant in scenarios demanding high precision (e.g., single-shot time of flight). Our findings demonstrate that cross-device variability in CSI is real but manageable, and we provide tools and guidelines for robust, hardware-agnostic Wi-Fi sensing.

This article presents a closed-loop, uncertainty-aware framework for defending a protected zone against coordinated incursions by swarms of small uncrewed aircraft systems (UAS). The interaction structure of the attackers is modeled as time-varying, while defenders operate under imperfect sensing. The proposed criticality-driven defender-to-attacker assignment strategy integrates three components: a probabilistic graph-based representation of the attacking swarm inferred from uncertain observations; a risk-aware attacker criticality model combining time-to-breach urgency with uncertainty; an online defender allocation mechanism that assigns and selectively reassigns defenders while limiting switching-induced instability through robust execution constraints. Analytical guarantees are established within a filtration-based first-hitting-time framework. In particular, finite-time triggering of the first capture event following detection is proven, and explicit mixed linear-geometric upper bounds are derived for the expected neutralization time. Monte Carlo simulations demonstrate the effectiveness of the proposed framework, achieving 85.6% neutralization efficiency under probabilistic sensing and 99.9% under deterministic sensing. Systematic ablation and sensitivity studies further quantify how detection thresholds and coordination parameters influence reliability and time-to-first-capture.

This paper presents novel algorithms for multi-target direction-of-arrival (DoA) estimation in array signal processing. Although the maximum likelihood estimator (MLE) asymptotically attains the Cramér-Rao bound, its exponential complexity motivates practical alternatives, such as greedy or subspace-based methods. In this context, greedy methods such as orthogonal matching pursuit (OMP) and orthogonal least squares (OLS) are sensitive to early selection errors, especially for angularly proximate targets, whereas subspace-based methods such as multiple signal classification (MUSIC) present angular super-resolution capabilities but degrade under strong inter-target signal correlation. To overcome these limitations, we propose two greedy iterative MUSIC (G-iMUSIC) algorithms, namely OMP-iMUSIC and OLS-iMUSIC, derived from a unified framework that links subspace and greedy estimations. Unlike prior iMUSIC approaches, the proposed methods require only one initial eigen value decomposition (EVD) and avoid computing eigendecomposition at each iteration. They also admit Fast Fourier Transform (FFT)-accelerated implementations for uniform linear arrays (ULAs), enabling low-complexity operation. Monte Carlo simulations demonstrate improved detection and precision over conventional OMP, OLS, and MUSIC, as well as reduced processing time compared to greedy baselines. Finally, we introduce diagnostic metrics that interpret performance across signal correlation and angular proximity regimes, supporting generalization beyond the specific orthogonal frequency-division multiplexing (OFDM) radar scenario considered.

Gaussian Splatting (GS) has emerged as an efficient representation for high-quality 3D reconstruction and novel view synthesis. However, its large model size poses challenges for storage and transmission. While several GS compression solutions have been proposed, their perceptual impact remains poorly understood due to the lack of dedicated evaluation datasets. To address this gap, this paper introduces GScomp-QA, a subjective quality assessment dataset for evaluating synthesis quality from compressed GS models. The dataset comprises 331 video stimuli from 13 real-world scenes, covering 9 state-of-the-art GS compression solutions. By using videos synthesized from uncompressed models as reference, GScomp-QA isolates compression-induced distortions from synthesis artifacts. A subjective study with 20 participants was conducted, providing reliable perceptual scores. Based on these data, GS compression solutions are evaluated through perceptual rate-distortion analysis. In addition, 18 objective quality metrics are evaluated, showing that they do not fully capture GS-specific distortions. GScomp-QA will be publicly available and provide a benchmark for evaluating GS compression solutions and supporting the development of quality metrics tailored to GS compression.

The growing adoption of electric vehicles (EVs) is increasing peak demand in distribution systems, which can threaten grid stability and reduce operational efficiency. Dynamic electricity pricing is a promising means of mitigating these peaks by shifting flexible demand. However, most existing approaches rely on detailed user-level consumption data and behavioral models, which are often difficult to obtain in practice and may raise privacy concerns. This paper proposes an Online Feedback Optimization (OFO) algorithm for day-ahead price design with limited data, where only aggregate loads are observed. OFO updates prices iteratively using aggregate load measurements, enabling effective peak reduction without access to individual user data. The formulation also includes a term that penalizes deviations in total electricity cost relative to a reference tariff. Although relying only on aggregate load measurements, the OFO price updates efficiently converge to the optimal price. In finite-horizon simulations, OFO achieves peak reduction close to that of the Stackelberg benchmark with full model information. Meanwhile, its computational effort is substantially lower. Additional tests under multiple initial conditions and delayed charging-window mismatch further confirm the robustness of the proposed method. Overall, these results show that OFO is a scalable and computationally efficient approach for peak-demand management in distribution systems with limited observability.

This paper introduces a Gaussian process (GP)-based method for extended object estimation (EOE) in integrated sensing and communication (ISAC) scenarios, representing a promising approach to enhance environmental awareness beyond the conventional point-scatterer assumption. The suitability of the proposed GP-based method for EOE is investigated through a practical measurement setup compliant with the fifth-generation (5G) New Radio (NR) standard and employing bistatic sensing, with results evaluated for both mapping and simultaneous localization and mapping (SLAM ) cases at millimeter-wave (mmWave) frequencies. The findings reveal that the enhanced capabilities of communication networks, when combined with bistatic sensing and GP-based EOE, enable improved environmental awareness in future wireless systems. Importantly, the results demonstrate that, under practical conditions, GP effectively performs EOE in both mmWave mapping and SLAM scenarios.

User localization and beam management are tightly linked in extremely large-scale multiple-input multiple-output (XL-MIMO) systems, especially in dense low-altitude economy (LAE) scenarios. However, the near-field propagation in XL-MIMO introduces strong distance sensitivity and complex spatial coupling, which makes joint trajectory and beam prediction challenging. Meanwhile, large language models (LLMs) have attracted attention in physical-layer transmission for modeling long-range dependencies. In this paper, we propose NF-TrackLLM, a multi-modal semantic-aware framework for near-field unmanned aerial vehicles (UAVs) positioning and beam prediction in XL-MIMO systems. By incorporating visual and LiDAR sensing into a Sionna-based channel generation pipeline, environmental semantics and GPS are utilized to guide trajectory and beam prediction. Built upon the aligned multi-modal representation, a GPT-2-based spatiotemporal reasoning backbone, and a cascaded prediction strategy are employed, where future trajectories are first inferred and then used to guide beam prediction as geometric priors. Simulation results demonstrate that NF-TrackLLM achieves accurate beam prediction and reliable UAV trajectory tracking in dense urban low-altitude scenarios.

Low Earth Orbit (LEO) satellite constellations are emerging as a key component of non-terrestrial networks due to their low-latency and high-capacity communication capabilities. However, satellites in these orbits are characterized by a small coverage footprint and high orbital velocity compared to those in higher orbits. This results in constantly changing and dynamic constellations that require smart design of orbital parameters to ensure continuous coverage. Existing constellation deployments are typically optimized either for low- and mid-latitude regions or for full polar coverage, leaving high-latitude regional scenarios such as the North Atlantic insufficiently explored. This work provides insights into the key characteristics associated with the deployment of satellites in LEO for North Atlantic coverage. Therefore, we investigate how constellation inclination, minimum elevation angle, altitude, and satellite footprint jointly affect visibility probability, revisit time, path loss, and coverage continuity. Results show that the minimum elevation angle is a critical design parameter since a Walker Delta constellation with 64 satellites at 1000 km altitude can provide continuous coverage above 55°N for elevations below 20°, whereas coverage probability degrades drastically for larger elevation angles. Similarly, inclinations above approximately 70° are required to achieve robust North Atlantic coverage with medium-size constellations. Thus, these results provide practical guidelines on how a satellite constellation should be designed to achieve an efficient deployment with a focus on coverage over the North Atlantic, targeting maritime, aviation, and Arctic connectivity scenarios.

In recent years, progress in adaptive graph signal processing algorithms has provided effective solutions for processing signals defined on graph structures. As a classical strategy in information theory, the Generalized Maximum Correntropy Criterion (GMCC) exhibits good resistance to non-Gaussian noises. When non-Gaussian noise interferes with the graph signal, the graph signal processing algorithm based on GMCC (GSP GMCC) algorithm shows better performance. However, the GSP GMCC algorithm itself has three parameters that need to be manually tuned, and the process of manually tuning the parameters is complex and tedious. Meanwhile, the non-concave and non-convex nature of the GMCC function itself limits its own convergence rate and adaptive estimation accuracy. To solve the above problems, based on the strongly convex function half-quadratic criterion (HQC), the GSP HQC algorithm is proposed in this paper. The performance analysis of the GSP HQC algorithm is implemented in this paper. Simulation experiments demonstrate that the GSP HQC algorithm achieves superior performance in terms of convergence rate and adaptive estimation accuracy while maintaining computational complexity comparable to existing algorithms

To support the adoption of electric transport systems, public charging opportunities are becoming increasingly important. In this dynamic environment, a central challenge for route planning and charging scheduling is forecasting charging-station availability under fluctuating demand. In this work, we propose a fluid-based forecasting method that accounts for uncertainty in both known and unforeseen electric vehicle arrival patterns while respecting station capacity constraints. We further evaluate the congestion forecasting method by applying it to an electric vehicle scheduling problem. Compared to scheduling frameworks that rely on standard baselines, charging schedules based on the fluid congestion forecasting model reduce waiting-related downtime by up to 14%. Finally, we quantify how increased knowledge of vehicle arrivals and different levels of station congestion affect overall system performance.

This paper presents a controlled over-the-air (OTA) characterization of dual-user IEEE 802.11be Extremely High Throughput Multi-User (EHT-MU) transmission under transmit-chain imbalance. The objective is to determine whether attenuation applied to one access-point transmit chain produces packet-global degradation or appears primarily as stream-dependent payload degradation after receiver processing. Measurements are performed in a shielded RF enclosure using two NI USRP-2953R and NI USRP-2942R software-defined radios, with one USRP generating a dual-user non-OFDMA EHT-MU waveform and the other implementing synchronized dual-branch packet recovery. A calibrated attenuation sweep is applied to the second AP transmit chain (TX2), and performance is evaluated using bit error rate (BER), EHT-Data error vector magnitude (EVM), control-field success probability, payload-success probability, and subcarrier-level EVM distributions. The results show that the stream decoded as User~1 remains at the BER floor over the tested range, while the stream decoded as User~2 exhibits progressive EVM degradation followed by threshold-like BER and payload-success collapse. Common signaling fields remain recoverable, indicating that the dominant observed failure mode is stream-local at the receiver output than the packet-global. Replacing User~2 binary convolutional coding (BCC) with low density parity check (LDPC) coding delays the BER and payload-success collapse by approximately \(5\)~dB of TX2 attenuation, demonstrating a measurable coding-dependent robustness margin for the more sensitive stream.

An over-the-air (OTA) experimental evaluation of concurrent 5G New Radio (5G NR) and Wi-Fi transmission using successive interference cancellation (SIC) in a shielded-box environment is presented. A USRP is used as the receiver, which captures the composite waveform containing both air-interface signals and applies sample-domain SIC to suppress the dominant 5G-NR signal and recover Wi-Fi signal from the residual waveform. The framework reports error vector magnitude (EVM), bit error rate (BER), sample-domain cancellation depth, and channel-estimate suppression, and, at the representative \(18\) dB attenuation point, measures \(11.88\) dB cancellation depth and \(26.96\) dB 5G channel suppression. The proposed methodology provides a practical basis for assessing cross-technology coexistence and receiver-side interference suppression under controlled OTA conditions.

Modern energy systems in vehicles and built infrastructure are governed by high-dimensional dynamics spanning multiple physical domains (e.g., electrical, thermal, mechanical) and timescales. This tutorial paper presents a graph-based modeling approach created to facilitate the modeling, analysis, control, estimation, optimization, and design of these systems. Matured and validated through more than a decade of research spanning multiple academic institutions and companies, the graph-based approach combines transient energy conservation with an explicit mathematical representation of the network by which energy is stored and transferred within a system. Following a mathematical overview of graph-based models, examples of multi-domain component and system models from the recent literature are presented, including single-phase thermal systems, two-phase thermal systems, and electro-mechanical systems. This is followed by a survey of recent applications for decentralized and hierarchical model predictive control, design optimization, and control co-design. Lastly, the paper describes an open-source toolbox created to facilitate the generation and analysis of graph-based models.

This paper studies how to balance onboard and ground computation under intermittent LEO connectivity for optimized inference freshness. As connectivity varies in time, the system switches among the actions of onboard computation, cached semantic transmission, raw-data offloading, and waiting. We define Age of Inference (AoInf) as the performance metric, where the age resets only upon successful task-valid updates. We formulate long-run average AoInf minimization as a finite-state average-cost semi-Markov decision process whose state captures the ground AoInf, orbital contact phase, cache occupancy, and cache age. We then transform the SMDP into an equivalent average-cost MDP and compute the solution via normalized relative value iteration (RVI). Numerical results indicate that the resulting hybrid policy reduces average AoInf relative to onboard-only and offload-only baselines, while requiring less computational resources on the satellite than the former, and fewer communication resources than the latter.

Large audio language models (LALMs) process both speech and environmental acoustic cues, yet struggle to retain non-speech information across multi-turn interactions. The performance gap between semantic (speech) and acoustic (non-speech) understanding remains poorly understood, and the underlying mechanisms of representation and retrieval are still unclear. This work introduces EnvMem, a controlled multi-turn benchmark designed to study this gap and identify the root causes of failures at the representation (i.e., latent embeddings) and retrieval levels (i.e., attention allocation). We further conduct post-hoc interventions to probe representational structure and attention dynamics. Our results reveal representational trajectory drift as the key failure mode, while showing that attention allocation plays a limited role in explaining the observed degradation. Overall, we provide a systematic framework for analyzing and improving non-linguistic memory in long-context LALMs, shedding light on future data and training design for robust acoustic memory modeling.

The electrification of long-haul freight transport introduces several new challenges, such as the limited capacity and congestion at en-route charging infrastructure. To reduce waiting times during peak periods, this paper proposes a framework for coordinated charging scheduling. The approach employs a mixed-integer formulation to optimize charging-related costs across charging, operation, battery degradation, and congestion delay, considering a range of scenarios. The results demonstrate that coordinated scheduling yields substantial cost savings up to 36% compared to uncoordinated scheduling, particularly by reducing battery degradation and delay costs.

Electron tomography (ET) plays an important role in the three-dimensional (3D) characterization of nanomaterials. However, under limited-angle and sparse-view conditions, conventional algorithms produce degraded reconstructions, which compromise the quality and interpretability of resulting 3D data. In this paper, we present deep image prior (DIP), an unsupervised deep learning (DL) approach, for highly degraded tomography acquisitions and demonstrate, using simulated data, that its performance is comparable to that of supervised approaches requiring training datasets, even for tilt ranges as limited as 60° and tilt increments of 10°. We then apply it to experimental data and show that it enables reliable 3D quantification under both sparse-view and limited-angle conditions, highlighting its potential for a wide range of materials and acquisition modalities.

Unloading containers in the courier, express and parcel industry is a physically demanding and labor-intensive work. Automatizing this process is an important step towards increasing the efficiency of parcel-handling systems. This work investigates the potential of reinforcement learning to learn a policy for item selection in container unloading scenarios. For that, a simulation environment is created and a masked deep Q-learning with a specially designed neural network architecture is implemented. The results indicate that the agent can learn to select items with an average success rate of 60 %, which is significantly better than a random policy at a random chance of 20 %. The findings suggest that RL could be a promising approach for automatizing item unloading tasks in the future.

Wireless digital twins require repeated synchronization between a time-evolving physical scene and its digital counterpart under limited and time-varying communication resources. For perception-centric twins, pixel-domain transmission or uniformly protected bitstreams can be mismatched to the semantic state consumed by twin-side applications. This paper proposes TWIST, a closed-loop token synchronization framework for application-aware wireless digital twins. TWIST represents each physical observation as a token and synchronizes this state over a wireless link, rather than optimizing visual reconstruction. Token positions are grouped by task relevance and protected through mode-conditioned unequal error protection under low-, medium-, and high-synchronization modes. At the twin side, decoding confidence converts unreliable hard token decisions into erasures, which are restored by a completion model before updating the semantic twin state. The recovered state supports traffic-state inference and generates compact feedback statistics, including channel quality, receiver uncertainty, semantic drift, and application priority, for subsequent mode adaptation. Experiments on a dynamic road-scene digital-twin scenario show that TWIST improves traffic-state inference and semantic twin-state synchronization compared with fixed-mode and channel-only adaptation strategies, while reducing the average synchronization cost relative to always-high transmission.

This paper tackles the optimization of the point spread function (PSF) of unmanned aerial vehicle (UAV)-borne multiple-input multiple-output (MIMO) synthetic aperture radar (SAR) tomography systems. A swarm of UAV-borne SAR systems is deployed to image an area to obtain its height profile. To achieve a high-quality three-dimensional (3D) image of the scene, the PSF has to exhibit low sidelobes. The heavy computations, required for image generation, are performed on the ground. To this end, the sensor data collected by the UAV-SARs is offloaded in real time via a frequency division multiple access (FDMA) air-to-ground backhaul link. In this work, the UAV formation and the power allocated for offloading are jointly optimized for the minimization of the PSF sidelobe levels. To this end, we propose a novel solution based on the particle swarm optimization (PSO) algorithm, which meets practical sensing and communication constraints. Our simulation results demonstrate that the proposed solution can significantly improve sidelobe suppression compared to several benchmark schemes.

Purpose: Real-time (RT) bSSFP MRI enables fast free-breathing cardiovascular imaging but requires 10-16 slices for functional assessment, resulting in prolonged scan times. Simultaneous multi-slice (SMS) imaging can reduce acquisition time but when combined with non-Cartesian trajectories, it relies on iterative reconstructions that preclude online use. This study investigates deep artifact suppression to facilitate rapid, online reconstruction of RT-SMS. Methods: A spiral bSSFP SMS RT sequence with two simultaneously acquired slices was implemented at 1.5 T. Reconstruction used slice separation in k-space, followed by deep artifact suppression in image space using a 3D U-Net. Ten healthy volunteers were imaged. RT-SMS image quality and reconstruction time were compared between deep artifact suppression and compressed sensing (CS) reconstructions. Left (LV) and right (RV) ventricular volumes at end diastole (EDV) and end systole (ESV) and LV mass (LVM) were compared between RT-SMS with deep artifact suppression and reference-standard breath-hold (BH) imaging. Results: The RT-SMS acquisition was ~13x faster than BH imaging (15 s vs 3 min 15 s). RT-SMS reconstruction using deep artifact suppression was ~50x faster than CS (30 s vs 24 min 55 s). Deep artifact suppression consistently outperformed CS in quantitative and qualitative image quality (p<0.001). Functional agreement between BH and RT-SMS with deep artifact suppression was good (LVEDV: -7.5 +/- 6.8 ml, LVESV: -0.9 +/- 4.2 ml, RVEDV: -6.4 +/- 8.4 ml, RVESV: 0.2 +/- 10.7 ml, LVM: -10.3 +/- 11.0 g). Conclusion: Online deep artifact suppression reconstruction for RT-SMS bSSFP CMR enables free-breathing short-axis coverage with a substantial reduction in acquisition and reconstruction time while maintaining diagnostic image quality.

Convolutional neural receivers such as DeepRx outperform minimum mean-square error physical uplink shared channel detection on in distribution channel and waveform configurations, but their behavior under calibration drift when transmitter or channel parameters depart from the training envelope is poorly characterized.

This paper presents the HRVConformer, a novel deep learning architecture for the classification of hypoxic-ischemic encephalopathy (HIE) using the instantaneous heart rate (HR) signal. Unlike conventional approaches that rely on handcrafted features, HRVConformer directly processes raw HR signals in an end-to-end manner, capturing both local and long-range dependencies through a hybrid Convolution-Transformer framework. By integrating convolutional layers for local feature extraction and Transformer-based attention mechanisms for global context modelling, the architecture effectively enhances signal representation and classification performance. The model was trained using supervised learning on a large HR dataset consisting of 1,573 one-hour epochs, including 259 one-hour expert-annotated epochs and a substantial set of weakly labelled data. A 314-hour validation set provided a robust performance estimation, while an independent 215-hour dataset with expert annotations was reserved for final testing. HR signals were extracted from electrocardiogram (ECG) recordings using an improved Pan-Tompkins algorithm, which significantly enhanced both signal quality and data availability. Experimental results demonstrate that the HRVConformer achieves an AUC of 83.23\% and accuracy of 74.56\% on the test set. These results surpass the performance of the Transformer, ResNet50 and fully convolutional networks baselines, highlighting the advantages of integrating convolutional and Transformer-based components for HR-based HIE classification. The proposed method provides a promising step toward a more accurate and automated assessment of HIE using HR signals. The code is available at: this https URL.

Chunk-wise autoregressive video diffusion models rely on a KV cache of previously generated chunks to avoid redundant computation, but this cache quickly becomes a memory bottleneck as videos grow longer. Methods that quantize the KV cache to low bitwidths reduce memory pressure but degrade video quality. We show that a key driver of this degradation is a systematic bias in attention weights: due to the convexity of the exponential in softmax attention, quantization noise inflates the contribution of cached keys, a phenomenon we call the Jensen bias. This effect causes quantized keys to steal attention mass from the unquantized current chunk. We derive a per-attention-score correction that removes this bias in expectation, computed on the fly from the quantization step sizes of the cached keys and the query norm. Using a second-order Taylor approximation, the additional computational overhead is negligible, and no additional memory is needed alongside the cache. Evaluated on MAGI-1, SkyReels-V2, and HY-WorldPlay at INT2 quantization, our correction recovers most of the quality lost to aggressive quantization, reaching near-BF16 video quality, and can outperform INT4 quantization while using 50% less memory.

This paper details two novel frameworks for developing autonomous, agentic AI in scientific workflows. Both systems leverage a hybrid Local Body, Remote Brain architecture via Google Colab, utilizing Python-based local orchestrators to invoke large language model (LLM) cloud backends. The first agent, DeepTS/DeepCollector, automates the large-scale curation, extraction, and deduplication of time-series datasets. The second, DeepScribe, is an autonomous presentation analyzer that converts visually dense, mathematically complex physics lectures into structured scientific reports. Through practical systems engineering-such as granular attribute extraction (Cellular RAG), remote data inspection, and distributed concurrency controls-we demonstrate how agentic AI can overcome the context and reasoning limitations of current state-of-the-art systems to rigorously support scientific workflows. Finally, we outline a generalization of DeepTS to support deep knowledge graphs and discuss the application of this conceptual approach to high-energy physics (DeepQCD).

Standard transformer attention computes pairwise token similarity but treats all tokens as equally salient and all positions as equally local, regardless of the informational structure of the input. We identify two complementary inductive biases that standard attention lacks: energy salience (which tokens concentrate informational energy, learned end-to-end without explicit frequency decomposition) and scale-selective locality (how far positional influence extends at each frequency, implemented via Morlet wavelet encoding). We address both with two simple components. Energy-Gated Attention (EGA) gates value aggregation by a learned energy estimate of key token embeddings, computed via a single linear projection; it selects what to attend to. Morlet Positional Encoding (MoPE) replaces fixed sinusoidal encodings with learned Gaussian-windowed wavelets that adapt the joint position-frequency localization to the corpus; it specifies where attention operates at each scale. On TinyShakespeare, EGA alone achieves +0.092 validation loss improvement over standard attention (+0.103 over Phase 1-3 baseline); MoPE alone is -0.032 (below baseline as a standalone encoding); but their combination achieves +0.119 -- more than the sum of parts. This superadditivity, observed across two independent training runs, is the central empirical finding: salience and locality are complementary inductive biases, each addressing a gap the other cannot fill alone. Ablations confirm that structured spectral priors (Morlet wavelet gates, scale-initialized heads, fixed sinusoidal PE) consistently underperform their unconstrained learned counterparts, while complementary learned components interact superadditively. All experiments are at small scale (<=6M parameters, character-level benchmarks, single seed); larger-scale multi-seed validation is the most important direction for future work.

At global scale, data-center electricity demand is growing faster than the grids that supply it, while system operators increasingly require large flexible loads that can adjust power within seconds to absorb variable wind and solar generation. For multi-megawatt AI/HPC facilities, the key unresolved question is practical and measurable: how quickly can the software stack translate a grid request into a real change in GPU power at the facility meter, where commitments are settled? We answer this on real hardware with GridPilot, a three-tier predictive controller operating across milliseconds, seconds, and hours, augmented by a deterministic safety-island bypass for fast response. On a three-GPU NVIDIA V100 testbed, GridPilot achieves a measured end-to-end trigger-to-target response of 97.2 ms, which is 6.9x faster than the 700 ms requirement of Nordic Fast Frequency Reserve. We further incorporate an instantaneous Power Usage Effectiveness (PUE) correction so dispatched commitments remain robust at meter level rather than only at IT load level. In replay experiments across six representative European grids (from Sweden to Poland), the PUE-aware controller closes 2.5-5.8 percentage points of cooling-overhead drag. GridPilot is released as open source and serves as a proof of concept that MW-scale AI/HPC demand can be engineered as controllable, grid-responsive flexibility by design.

Task-based fMRI provides a direct readout of task-evoked neural dynamics, but it is expensive and difficult to acquire at scale, motivating rest-to-task synthesis from widely available resting-state fMRI (rsfMRI). We propose FM-fMRI, an event-conditioned flow-matching model that learns a continuous-time conditional vector field to generate task ROI time series from a subject's rsfMRI and the task event information. The formulation enables fast ODE-based sampling and flexible conditioning over heterogeneous event schedules. Rather than optimizing for pointwise reconstruction, we evaluated generated signals using complementary criteria that probe temporal and spectral structure, subject and group-level connectome consistency, and distributional alignment. On the public Human Connectome Project and internal BioPoint autism cohort, FM-fMRI achieves the strongest spectral and connectivity agreement and improved distribution-level matching over conditional diffusion, generative adversarial networks (GANs), and variational autoencoders (VAEs) baselines. Furthermore, we augment the BioPoint cohort by synthesizing task-fMRI ROI time series with our method, improving downstream autism classification and demonstrating practical utility in data-limited clinical settings. The code will be available on GitHub.

Safe reinforcement learning (RL) for robotic systems requires policies that improve task performance while satisfying state and input constraints during both training and deployment. Control barrier functions (CBFs) provide a principled mechanism for enforcing forward invariance through minimally invasive safety filters, but their use in model-free RL is limited by the need for accurate dynamics and hand-designed barrier certificates. We propose Robust Koopman-CBF SAC, a safety-filtered actor--critic framework that learns a finite-dimensional Koopman predictor from data, constructs affine CBF constraints in the lifted space, and enforces them through a quadratic-program safety layer. To account for finite-dimensional Koopman approximation error, the CBF condition is tightened using a projected residual margin estimated from held-out rollout data. The critic is trained on the executed safe action, while the actor is regularized toward the Koopman-CBF feasible set, reducing dependence on the filter over training. Across safe-control benchmarks, the method achieves zero constraint violations on CartPole stabilization and tracking while matching or exceeding unconstrained SAC returns. On high-dimensional Safety Gymnasium locomotion tasks, the method reduces violations in some settings but also exposes important limitations of first-order velocity barriers and linear EDMD models, motivating high-order and multi-step Koopman-CBF extensions. These results suggest that robust Koopman-CBF filters are a promising bridge between model-free RL and certifiable safety, while clarifying the structural conditions under which such filters remain effective. All code is available at \href{this https URL}{Github Repository}.

We present the stochastic decoupled policy gradient (SDPG), a lightweight visual reinforcement learning (RL) method that trains diverse visuomotor control policies end-to-end within a few hours on a single NVIDIA RTX 4080 GPU. SDPG estimates policy gradients via random perturbations of trajectory rollouts, requiring orders of magnitude fewer batch-rendered environments and substantially reducing compute and memory overhead. On visual MuJoCo benchmarks, SDPG consistently outperforms baseline methods in training time, memory usage, and rewards. Finally, to support future research, we introduce a suite of realistic visual robotics benchmarks spanning dexterous manipulation, challenging locomotion, and demonstrate effective sim-to-real transfer on physical hardware.

With the widespread application of location-based services, fingerprint-based localization has demonstrated advantages in environments with complex signal propagation. Deep learning has significantly improved the efficiency of both offline training and online matching in localization processes. However, existing fingerprints only contain terminal position information without capturing motion states, and neural network designs have not fully incorporated structural features such as fingerprint sparsity. In this paper, we propose a triple-beam fingerprint (TBF) incorporating Doppler information and design a Transformer-based localization and orientation awareness network (LOA-Net) to simultaneously estimate user position and motion direction in massive multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems. We first show the correlation between TBF and multipath information, and investigate the collinearity of different TBFs, demonstrating that TBF is an effective small-size sparse fingerprint. Then, we propose LOA-Net containing a mask-augmented detection Transformer for regression (MaskDETR-Reg) module and a fusion-enhanced Transformer for direction classification (Fusion-TDC) module to process angle-delay domain information and Doppler domain information, respectively. Finally, in the simulation of indoor scenarios defined in 3GPP 38.901, the proposed method achieves significantly better localization accuracy than weighted $K$-nearest neighbors (WKNN), 2D and 3D convolutional neural networks (CNNs), and achieves satisfactory motion direction estimation accuracy.

In joint multiuser decoding, a receiver recovers a set of messages from a single noisy aggregate of many simultaneous transmissions. Classical decoders rely on rule-based mechanisms such as successive interference cancellation, joint belief propagation, or list recovery, all of which become brittle or expensive as ambiguity increases. We propose CIDER, a learned multiuser decoder with masked-diffusion refinement steps. CIDER uses demixing to prevent duplicate-row collapse and uses parity-aware propagation to provide soft guidance from the code constraints. In higher-load regimes, we further improve reliability via a lightweight quality-guided remasking step that selectively re-decodes low-confidence sequences. On commonly used error-correcting codes, CIDER matches or improves on FFT-accelerated joint belief propagation-style decoding in symbol error rate while running more than $6\times$ to over $100\times$ faster, with the speedup widening as the blocklength grows. Code is available at this https URL.

We present a provably safe sampling-based motion planning algorithm for robotic systems affected by random disturbances of unknown distribution. We consider systems with linear or linearizable dynamics evolving in workspace with arbitrary-shaped obstacles subject to state and control constraints. Safety requirements are formulated as chance-constraints. Our approach leverages data from trajectories of the system to learn a Wasserstein ambiguity tube, i.e., a sequence of ambiguity sets, which contains the trajectory of the system's state distribution with high confidence. This ambiguity tube is then used in a probabilistically complete algorithm to grow a sampling-based motion planning tree that respects the constraints of the problem. We show that learning several lower-dimensional ambiguity tubes instead of a single high-dimensional one effectively reduces the conservatism and boosts scalability. Additionally, we design an efficient bandit-based validity checker that remarkably increases the empirical performance of our approach without sacrificing probabilistic completeness. Case studies show our algorithm finds valid plans in cluttered environments under strict safety thresholds, outperforming state-of-the-art methods.

Pinching antenna systems (PASS) enable reconfigurable radiating elements and extended line-of-sight communication, mitigating path loss effects. However, existing designs lack fully controllable radiation weights, as they are governed by structural parameters rather than explicitly assigned variables. In this paper, we introduce a new degree of freedom (DoF) for PASS by enabling radiation weight control through phase-mismatch manipulation of guided waves under single-mode excitation within a coupled-mode framework. By tuning the propagation constants of pinching antennas, independent complex-weight control of individual elements is achieved, transforming PASS into a weight-adaptive analog beamforming architecture. Based on this principle, we present a physics-based hardware model that provides a unified framework for both amplitude-tunable pinching beamforming and conventional equal-power radiation models, ensuring compatibility with existing PASS implementations, such as movable setups. To evaluate the proposed model, we formulate a sum-rate maximization problem for hybrid precoding in multiuser downlink systems and solve it using an alternating optimization framework that combines weighted minimum mean square error-based digital precoding with genetic algorithm-based optimization of PASS configurations, including various scenarios such as weight tuning, antenna movability, and discrete activation. Numerical results demonstrate that the amplitude-tunable PASS architecture achieves consistent performance gains over conventional arrays and existing PASS schemes, with pronounced improvements in interference-limited regimes under practical constraints.

The increasing densification of small-cell networks substantially expands cable-based backhaul infrastructure, creating heightened vulnerability to cable link failures. This paper proposes a reconfigurable intelligent surface (RIS)-assisted backup framework that exploits a key insight: during backhaul cable failures, base station (BS) radio components remain functional, enabling wireless backhaul traffic redistribution. Our framework maintains network connectivity by redistributing disconnected BS backhaul traffic to neighboring BSs through RIS-assisted wireless links. To maximize survivability across varying traffic conditions, we formulate a joint optimization problem that maximizes total resolvable backhaul traffic by jointly deciding BS selection, RIS phase shifts, and precoding vectors. The inherent non-convexity arising from coupling and quadratic fractional term is addressed through an alternating optimization algorithm that iteratively solves tractable convex subproblems via quadratic transformation. Comprehensive numerical evaluations demonstrate that the proposed RIS-enhanced framework significantly improves survivability from 58% to 72% under challenging high-intensity hotspot traffic conditions. Moreover, RIS provides the greatest gains for antenna-constrained systems by extending coverage to access more spare capacity of the distant BSs as well as enhancing the signal strength. Consequently, high survivability is achieved even with only two antennas per BS under moderate traffic intensity.

In this paper we propose dynamic output-feedback controller synthesis methods for discrete-time linear time-invariant systems. The synthesis goal is either to achieve dissipativity with respect to a given quadratic supply rate, or to achieve given $H_2$ performance level. It is assumed that the autoregressive model of system dynamics is unknown, expect for the noisy disturbance term which is not part of the performance channel. Instead, we have a recorded trajectory of inputs and outputs which can be corrupted by an unknown but bounded disturbance. Methods are formulated in terms of linear matrix inequalities parametrized by a scalar variable, while in noiseless case they reduce to linear matrix inequalities. Within the considered setting, synthesis procedures are non-conservative.

This paper develops a robust fixed time optimization framework for constrained problems that guarantees exact constraint satisfaction and convergence to KKT points within fixed time , independent of initial conditions. The approach treats the Lagrange multipliers as control inputs, composed of an equivalent control and a switching control, with the system states representing the decision variables. An equivalent control steers the gradient flow to a local KKT point asymptotically for nonconvex objectives and to unique global optimum in fixed time for convex objectives. Constraint enforcement is achieved by embedding the equality constraints directly as a sliding manifold, with a fixed time switching control ensuring rapid and reliable feasibility. The framework further accounts for the matched disturbances, providing robustness guarantees that are theoretically characterized and illustrated using spherical constraints. Numerical studies on a 3-bus AC optimal power flow problem and distributed consensus=based parameter estimation problem demonstrate the effectiveness, scalability and robustness of proposed approach.

Submarine eruptions, accounting for over 80% of Earth's volcanic activity, primarily occur along mid-ocean ridges, where shallow magmatic systems are accessible to high-resolution imaging. Yet, their remoteness often leaves them undetected. Recent seismic studies at the East Pacific Rise (EPR) 9°50'N-one of the most dynamic ridge segments, imaged the detailed architecture of the shallowest magma lens, but no data-constrained model yet explains how magma accumulates, migrates, or triggers eruptions. Similarly, the formation of oceanic crust remains poorly understood. While 2-D seismic data reveal only a few vertically stacked, transient magma lenses, our study applies matrix imaging, a novel technique in controlled-source seismology, to map the inner structure of on- and off-axis magma reservoirs. We uncover a conical on-axis reservoir and interconnected magma-rich zones throughout the crust. Combined with ophiolite evidence, these findings reveal that magma channels dominate the first 3 km for lower crust formation, while in situ crystallization prevails in the final 1 km, resolving a long-standing debate.

This study examines the relationship between speech representations and the hierarchical structure of cognitive assessment in mild cognitive impairment. Utilizing 5,754 German neuropsychological assessment recordings, we evaluate six cognitive tasks across three score levels: task, domain, and global levels. We compare hand-crafted acoustic features with self-supervised learning (SSL) embeddings. Results show that although SSL representations generally outperform hand-crafted features at lower levels, this trend reverses for MCI classification. Furthermore, task-specific constraints influence performance: tasks with greater response freedom exhibit performance dilution as hierarchical levels increase, suggesting ``specialist'' representations, whereas the performance of highly structured tasks increases toward higher levels, suggesting ``generalist'' representations. These findings show links between task constraints and assessment hierarchy in automated clinical speech analysis.

Quantum state tomography (QST) is a fundamental task in quantum information science that aims to reconstruct unknown quantum states from measurement data. However, the exponential growth of Hilbert-space dimension with system size makes full tomography of general quantum states statistically and computationally prohibitive. This challenge has motivated extensive research on structured quantum state tomography, where prior structure, such as low-rankness, tensor-network representations, shallow quantum circuits, and neural quantum states, can substantially reduce the effective degrees of freedom and enable scalable recovery. In this review, we provide a unified perspective on QST for structured quantum states through three closely related themes: compact state representations, measurement design, and computational algorithms. After reviewing common models for structured quantum states, we survey existing work on geometric preservation properties of measurement frameworks, ranging from informationally complete POVMs to randomized measurements, and their implications for sample complexity. On the algorithmic side, we review optimization methods for reconstructing structured quantum states from empirical measurements. By connecting QST with broader principles from compressive sensing, matrix sensing, and structured inverse problems, this survey highlights common theoretical foundations underlying sample complexity, measurement efficiency, and scalable recovery.

Modern intrusion detection systems generate thousands of alerts daily, but alert fatigue severely limits security operations effectiveness due to too many false positives or low-impact events. We address this by proposing a principled framework for alert prioritization based on subnormal Gaussian fuzzy numbers, explicitly modeling three sources of uncertainty: threat severity, detection confidence, and organizational risk attitude. Each alert is represented as a fuzzy number with the core indicating severity, spread indicating uncertainty, and height reflecting detection reliability. We apply ranking indices to prioritize alerts, allowing organizations to tune security posture through a risk-attitude parameter. Experimental validation on CIC-IDS2017 and NSL-KDD demonstrates greater robustness than baselines under detector degradation (0.9963 vs 0.8215 NDCGrel@100), with distinct differentiation in mid-confidence alerts and near-parity with baselines under robust detectors. The framework is theoretically grounded, computationally efficient, provides interpretable reasoning, and remains robust across detector families and miscalibration scenarios.

Reactive control is often considered insufficient for multi-objective tasks because conflicting objectives give rise to local minima. We argue this limitation is not inherent but arises from static encodings that fail to reflect how objectives currently interact. We exploit the interaction structure encoded in a graph-based world model by extending it with nullspace projections: conflicts are resolved where they arise by projecting lower-priority gradients into the nullspace of higher-priority ones, with priorities determined continuously from the current state. We demonstrate this in two domains where conflicts between objectives are central: navigation around non-convex obstacles, where static potential fields fundamentally fail, and planar pushing of non-convex objects, where our method achieves $100\%$ success across one-hundred configurations versus $0\%$ for the steepest-descent baseline and ${\sim}55\%$ for diffusion policy, without demonstrations or retraining. The same formulation transfers directly to a real robot with additional perceptual and kinematic constraints, accommodating them through the same mechanism.

Modern retrieval agents expose many configuration choices -- LLM, retriever, number of documents, number of hops, and synthesis strategy -- each shaping both answer quality and serving cost. Today, these pipelines are typically hand-tuned once per workload, leaving substantial per-query optimization untapped. We formulate the problem: given a natural-language query and either an accuracy or a budget target, select from a predefined pipeline catalog the configuration that minimizes cost or maximizes accuracy at inference time. We propose **BRANE**, which uses an LLM to convert each query into workload-specific characteristics, then trains a lightweight per-configuration predictor that estimates whether the pipeline will answer the query correctly. At inference time, **BRANE** selects the configuration that maximizes predicted correctness penalized by cost, exposing a tunable cost-quality tradeoff without retraining. Across MuSiQue, BrowseComp-Plus, and FinanceBench, **BRANE** consistently pushes the cost-quality Pareto frontier, matches the best fixed configuration's accuracy at up to 89% lower cost, and outperforms LLM-routing, rule-based, and fine-tuned Qwen3-4B baselines. These results show that per-query configuration of the full retrieval pipeline is a practical alternative to static workload-level tuning.

In many image processing applications (e.g. computational anatomy) a groupwise registration is performed on a sample of images and a template image is simultaneously generated. From the template alone it is in general unclear to which extent the registered images are still spatially misaligned. The local spatial variation left after registration may be seen as the resolution of the resulting template. In this sense we develop the first template resolution measure (TRM) quantifying the misalignment at each location of the template. The TRM is based on the key insight that the size of such misalignments can be determined as the amount of smoothing required to bring the registered images in agreement. This relationship is mathematically derived from characteristic examples in one dimension. This way, the TRM quantifies the remaining spatial variation in the template's units of length. Furthermore we propose to enhance the template by an effective visualization of its resolution measure. Finally we demonstrate the TRM's applicability and validate its interpretability for example datasets in two and three dimensions. The corresponding code is publicly available on GitHub.

In this paper, we aim to mitigate congestion in traffic management systems by guiding travelers along system-optimal (SO) routes. However, we recognize that most theoretical approaches assume perfect driver compliance, which often does not reflect reality, as drivers tend to deviate from recommendations to fulfill their personal objectives. Therefore, we propose a route recommendation framework that explicitly learns partial driver compliance and optimizes traffic flow under realistic adherence. We first compute an SO edge flow through flow optimization techniques. Next, we train a compliance model based on historical driver decisions to capture individual responses to our recommendations. Finally, we formulate a stochastic optimization problem that minimizes the gap between the target SO flow and the realized flow under conditions of imperfect adherence. Our simulations conducted on a grid network reveal that our approach significantly reduces travel time compared to baseline strategies, demonstrating the practical advantage of incorporating learned compliance into traffic management.

Accurate estimation of the relative attitude and angular velocity between two rigid bodies is fundamental in aerospace applications such as spacecraft rendezvous and docking. In these scenarios, a chaser vehicle must determine the orientation and angular velocity of a target object using onboard sensors. This work addresses the challenge of designing an Equivariant Filter (EqF) that can reliably estimate both the relative attitude and the target angular velocity using noisy observations of two known, non-collinear vectors fixed in the target frame. To derive the EqF, a symmetry for the system is proposed and an equivariant lift onto the symmetry group is calculated. Observability and convergence properties are analyzed. Simulations demonstrate the filter's performance, with Monte Carlo runs yielding statistically significant results. The impact of low-rate measurements is also examined and a strategy to mitigate this effect is proposed. Experimental results, using fiducial markers and both conventional and event cameras for measurement acquisition, further validate the approach, confirming its effectiveness in a realistic setting.

An important limitation of Inverter-Based Resources (IBR) is their reduced contribution to Short-Circuit Current (SCC), as compared to that of Synchronous Generators (SGs). With increasing penetration of IBR in most power systems, the reducing SCC poses challenges to a secure system operation, as line protections may not trip when required. In order to address this issue, the SCC ancillary service could be procured via an economic mechanism, aiming at securing adequate SCC on all buses. However, the suitability of markets for SCC services is not well understood, given that these could be prone to market power issues: since the SCC contributions from various SGs to a certain bus are determined by the electrical topology of the grid, this is a highly local service. It is necessary to understand if SGs at advantageous electrical locations could exert market power and, if so, how it could be mitigated. In order to fill this gap, this paper, for the first time, adopts an SCC-constrained bilevel model to investigate strategic behaviors of SGs. To address the non-convexity due to unit commitment variables, the model is restructured through a primal-dual formulation. Based on a modified IEEE 30-bus system, cases with strategic SGs placed at different buses are analyzed. These studies demonstrate that strategic agents exerting market power by manipulating service prices and extending operating periods could achieve up to triple revenues from SCC provision, which reduces market efficiency and would increase the financial burden on consumers. These findings highlight the need for careful market design, for which potential measures to mitigate these market power issues are also discussed.

A great amount of endeavor has recently been devoted to activity detection for massive machine-type communications in cell-free multiple-input multiple-output (MIMO) systems. However, as the number of antennas at the access points (APs) increases, the Rayleigh distance that separates the near-field and far-field regions also expands, rendering the conventional assumption of far-field propagation alone impractical. To address this challenge, this paper establishes a covariance-based formulation that can effectively capture the statistical property of hybrid near-far field channels. Based on this formulation, we theoretically reveal that increasing the proportion of near-field channels enhances the detection performance. Furthermore, we propose a distributed algorithm, where each AP performs local activity detection and only exchanges the detection results to the central processing unit, thus significantly reducing the computational complexity and the communication overhead. Not only with convergence guarantee, the proposed algorithm is unified in the sense that it can handle single-cell or cell-free systems with either near-field or far-field devices as special cases. Simulation results validate the theoretical analyses and demonstrate the superior performance of the proposed approach compared with existing methods.

Maintaining high energy efficiency (EE) in wireless networks is crucial, particularly with the adoption of massive MIMO technology. This work introduces a resource allocation framework that jointly optimizes transmit power assigned to each user and the number of active antennas, while explicitly accounting for a nonlinear Power Amplifier (PA). We consider a downlink MU-MIMO-OFDM transmission with zero forcing (ZF) precoding, Rayleigh fading channels, and soft-limiter PAs, with both ideal and realistic PA architectures. In contrast to existing formulations, our optimization framework avoids imposing an explicit transmit power constraint, since the nonlinear distortion inherently limits the feasible operating region. To solve the resulting non-convex problem, an alternating optimization approach is adopted that, by exploiting properties of the EE function, guarantees convergence to a stationary point. Extensive simulations demonstrate consistent performance gains over distortion-neglecting and power-only optimized baselines. In a scenario of a 5 km radius cell serving 60 randomly distributed users, the median EE gains over the distortion-neglecting allocation reach 40% for ideal PAs and 20% for Class B PAs, confirming high impact of the proposed solution.

We study deep state-space models (Deep SSMs) that contain linear quadratic-output (LQO) systems as internal blocks and present a compression method with a provable output error guarantee. We first derive an upper bound on the output error between two Deep SSMs and show that the bound can be expressed in terms of the $h^2$-error norms between the layerwise LQO systems. In particular, we show that reducing the $h^2$ approximation errors of the LQO systems placed in shallow layers is effective in reducing the derived upper bound on the output error. Next, we formulate an optimization problem for the derived upper bound and develop a gradient-based MOR method. In the numerical experiments, using the IMDb task from the LRA benchmark, we demonstrate the effectiveness of the proposed upper-bound-based compression method. In particular, we show that the number of trainable parameters can be reduced by approximately 60\% without retraining while maintaining the performance of the original model.

This paper studies group target trajectory intent as the outcome of a cooperative game where the complex-spatio trajectories are modeled using an NLP-based generative model. In our framework, the group intent is specified by the characteristic function of a cooperative game, and allocations for players in the cooperative game are specified by either the core, the Shapley value, or the nucleolus. The resulting allocations induce probability distributions that govern the coordinated spatio-temporal trajectories of the targets that reflect the group's underlying intent. We address two key questions: (1) How can the intent of a group trajectory be optimally formalized as the characteristic function of a cooperative game? (2) How can such intent be inferred from noisy observations of the targets? To answer the first question, we introduce a Fisher-information-based characteristic function of the cooperative game, which yields probability distributions that generate coordinated spatio-temporal patterns. As a generative model for these patterns, we develop an NLP-based generative model built on formal grammar, enabling the creation of realistic multi-target trajectory data. To answer the second question, we train a Graph Transformer Neural Network (GTNN) to infer group trajectory intent-expressed as the characteristic function of the cooperative game-from observational data with high accuracy. The self-attention function of the GTNN depends on the track estimates. Thus, the formulation and algorithms provide a multi-layer approach that spans target tracking (Bayesian signal processing) and the GTNN (for group intent inference).

This paper studies the estimation of outage probability in GSC/MRC SIMO systems under Rician fading in the rare-event regime. The difficulty arises from the evaluation of the CDF of a partial sum of ordered non-central chi-square random variables, motivating the use of enhanced Monte-Carlo methods. For independent fading, we propose partition importance sampling (PIS), a theory-driven estimator tailored to this structure, and prove that it achieves bounded relative error (BRE). We further adapt exponential twisting, proving its BRE property, and cross-entropy to this setting. We then extend ET and CE to correlated Rician fading, where the joint distribution of the power gains is no longer tractable, yielding the ETC and CEC estimators; ETC enjoys the bounded-relative-error guarantee under arbitrary mean and arbitrary covariance. Numerical experiments compare these methods with universal importance sampling and multilevel splitting for independent fading, and an asymptotic approximation in the correlated case. Empirically, CE shows the most robust performance in the independent case; PIS and ET are competitive but degrade for larger means, with ET further degrading when the selected subset is much smaller than the antenna array. ETC yields a better estimate than the asymptotic approximation for moderately rare events.

The aggregate flexibility region of distributed energy resources (DERs) quantifies the aggregate power shaping capabilities of DERs. It characterizes the distribution network's potential for wholesale market participation and grid service provision at the transmission level. To enhance flexibility and fully exploit the potential of DERs, this paper proposes a method to optimize the aggregate flexibility region through distribution network reconfiguration. First, we formulate the ellipsoidal aggregate flexibility region characterization problem as a two-stage adaptive robust optimization problem and derive an exact convex reformulation with a large number of second-order cone constraints. By exploiting the problem structure, we propose a scalable Benders decomposition algorithm with provable finite convergence to the optimal solution. Finally, we propose an optimal reconfiguration problem for aggregate flexibility region optimization and solve it using the custom Benders decomposition. Numerical simulations on the IEEE 123-bus test feeder demonstrate that, compared to existing approaches, substantial improvements in the aggregate flexibility region can be achieved over multiple scenarios with the optimized topology.

We study distributed control of networked systems through reinforcement learning, where neural policies must be simultaneously scalable, expressive and stabilizing. We introduce a policy parameterization that embeds Graph Neural Networks (GNNs) into a Youla-like magnitude-direction parameterization, yielding distributed stochastic controllers that guarantee network-level closed-loop stability by design. The magnitude is implemented as a stable operator consisting of a GNN acting on disturbance feedback, while the direction is a GNN acting on local observations. We prove robustness of the policy to perturbations in both the graph topology and model parameters. Numerical experiments validate the effectiveness of the proposed approach.

Robust output regulation for linear time-varying systems has remained an open problem for decades. To address this, we propose the trajectory-matching system immersion framework, by reformulating the regulator equation into a more insightful form. This perspective demonstrates that finding an internal model is equivalent to reproducing the steady-state output trajectories of a given forced system by constructing an unforced one. This reveals the fundamental influence of parametric uncertainties, yielding the precise algebraic boundary for robust regulation, termed finite linear parameterization. With this, we further demonstrate that uncertainties in time-varying systems can easily excite infinite-dimensional families of functions, making it impossible for a finite-dimensional regulator to achieve exact robust regulation. Hence, we establish a comprehensive approximate robust design, which yields a bounded tracking error that can be arbitrarily small, and avoids explicitly solving the regulator equation. Additionally, it can ensure exact regulation when the uncertainty influences the system in some specified ways. Overall, these results provide a general, executable framework for constructing an internal model-based design, and simplify the robust control implementation process.

Large Language Models (LLMs) have demonstrated strong semantic reasoning across multimodal domains. However, their integration with graph-based models of brain connectivity remains limited. In addition, most existing fMRI analysis methods rely on static Functional Connectivity (FC) representations, which obscure transient neural dynamics critical for neurodevelopmental disorders such as autism. Recent state-space approaches, including Mamba, model temporal structure efficiently, but are typically used as standalone feature extractors without explicit high-level reasoning. We propose NeuroMambaLLM, an end-to-end framework that integrates dynamic latent graph learning and selective state-space temporal modelling with LLMs. The proposed method learns the functional connectivity dynamically from raw Blood-Oxygen-Level-Dependent (BOLD) time series, replacing fixed correlation graphs with adaptive latent connectivity while suppressing motion-related artifacts and capturing long-range temporal dependencies. The resulting dynamic brain representations are projected into the embedding space of an LLM model, where the base language model remains frozen and lightweight low-rank adaptation (LoRA) modules are trained for parameter-efficient alignment. This design enables the LLM to perform both diagnostic classification and language-based reasoning, allowing it to analyze dynamic fMRI patterns and generate clinically meaningful textual reports.

We study how the vector-field structure of nonlinear port-Hamiltonian systems is reflected in the infinitesimal Koopman generator. The generator admits a natural bracket decomposition into a conservative interconnection-bracket derivation, a dissipative metric-bracket derivation, and an input-port derivation. The conservative component is formally skew-adjoint on a test space whenever the conservative flow preserves the reference measure and the relevant boundary terms vanish. The dissipative component is not claimed to be a positive operator on arbitrary observables; rather, the positive semidefinite object is the metric bracket $[f,f]_R=\nabla f^\intercal\mathbf{R}\nabla f\ge 0$, which yields the exact port-Hamiltonian energy balance for the Hamiltonian observable: \[ \mathcal{K}_{\mathbf{u}}\mathcal{H} =-[\mathcal{H},\mathcal{H}]_R +\mathbf{y}^\intercal\mathbf{u} \le \mathbf{y}^\intercal\mathbf{u}. \] We use these bracket identities to motivate finite-dimensional weak Galerkin and data-driven lifted models: when the Galerkin measure is conservative for the Hamiltonian interconnection flow and boundary terms vanish, the conservative contribution is skew in the Galerkin mass metric, while the dissipative bracket induces a positive semidefinite Dirichlet matrix. These identities motivate structure-preserving lifted port-Hamiltonian surrogates that are passive and support damping injection in the lifted coordinates, while distinguishing exact bracket identities, projection residuals, finite-data estimation error, and the residual and injectivity assumptions needed to transfer lifted conclusions back to the original nonlinear state.

The core of time series analysis lies in effectively modeling the physical laws within complex signals. Existing Transformer and Convolution Neural Network (CNN) architectures are often constrained by insufficient temporal inductive bias, restricted frequency extraction capabilities, or weak local phase alignment. To this end, this paper proposes Adaptive Network Based on Cascaded Harmonic Offset Routing (ANCHOR), an Adaptive Network based on Cascaded Harmonic Offset Routing. The model utilizes the Real Fast Fourier Transform (RFFT) to extract explicit dominant periods, injecting them as physical anchors into the dilation operators of multi-branch deformable convolutions. This guides the adaptive optimization of sampling locations in the time domain, achieving synergistic modeling of macroscopic periodic priors and microscopic geometric deformations. Furthermore, to address the quantization errors and picket-fence effects introduced by the discrete RFFT, this paper imports a continuously differentiable 1D Gaussian Radial Basis Function interpolation operator to replace traditional linear interpolation. This maintains the differentiability of the interpolation process and enhances the accuracy of sub-pixel phase compensation. Additionally, ANCHOR introduces an asymmetric routing mechanism and orthogonal channel partitioning to dynamically balance the extraction weights between high-energy strong signals and low-energy weak features. Multi-task benchmark experiments demonstrate that ANCHOR achieves the best or solid performance in short-term forecasting, anomaly detection, and time series classification tasks. Code is available at this https URL

Smartphone cameras face fundamental form-factor constraints that limit their optical magnification, primarily due to the difficulty of reducing a lens assembly's telephoto ratio, the ratio between total track length (TTL) and effective focal length (EFL). Currently, conventional refractive optics struggle to achieve a telephoto ratio below 0.5 without requiring multiple bulky elements to correct optical aberrations. In this paper, we introduce MetaTele, a novel optics-algorithm co-design that breaks this bottleneck. MetaTele explicitly decouples the acquisition of scene structure and color information. First, it utilizes a compact refractive-metasurface optical assembly to capture a fine-detail structure image under a narrow wavelength band, inherently avoiding severe chromatic aberrations. Second, it captures a broadband color cue using the same optics; although this cue is heavily corrupted by chromatic aberrations, it retains sufficient spectral information to guide post-processing. We then employ a custom one-step diffusion model to computationally fuse these two raw measurements, successfully colorizing the structure image while correcting for system aberrations. We demonstrate a MetaTele prototype, achieving an unprecedented telephoto ratio of 0.44 with a TTL of just 13 mm for RGB imaging, paving the way for DSLR-level telephoto capabilities within smartphone form factors.

Distributed massive MIMO (D-MIMO) is a promising technology for future generation wireless systems as it takes advantage of both an increased array aperture and a decentralized processing architecture and topology. In order to truly understand the possibilities and limitations of these approaches in real scenarios, practical realization of testbeds is an essential step in the technology advancement. This work presents the Lund University Large Intelligent Surface testbed -- LuLIS, that can operate up to 256 coherent radio frequency (RF) chains using 16 AMD Zynq UltraScale RFSoC ZCU216 evaluation boards acting as distributed processing nodes. Real-time processing is facilitated by acceleration and distribution of MIMO processing algorithms on the FPGA fabric of the boards. The system is easily scalable, as increasing the number of antennas is done in multiples of 16 by adding more RFSoCs, which also implies addition of another processing node. The design allows up-scaling without hardware redesign, introduction of large latencies or data transfer overhead. The testbed is flexible in terms of deployment, with options of fully distributing the nodes (as in D-MIMO) or co-locating them (as in more traditional Massive MIMO). A detailed description of the implementation of the testbed is presented and initial results are shown for an uplink (UL) transmission from four single-antenna user equipments (UEs) to 64, 128 and 256 base-station antennas.

In this white paper, we summarize for the benefit of the wider research community on wireless communications, the two key results that we shared with the attendees of the 2026 IEEE Communication Theory Workshop in Azores, Portugal, about affine frequency division multiplexing (AFDM). Firstly, we show that in contrast to the wide perception by most researchers, AFDM can be implemented at marginal costs by means of a simple software upgrade (firmware patch) of conventional orthogonal frequency division multiplexing (OFDM), indicating that its adoption can potentially be achieved across a wide range of OFDM-based wireless infrastructure and systems. The most crucial relevance of this finding is that such an upgrade would enable, under the specific conditions of the corresponding systems and their applications, exploiting various advantageous features of AFDM, including robustness to doubly dispersive channels (i.e., to support high-mobility use-cases in 6G), inherent integrated sensing and communications (ISAC) compatibility (i.e., to support sensing use-cases in 802.11bf), and the straightforward introduction of low-complexity physical-layer security at the waveform level (as needed in next-generation IoT systems). Secondly, we also show that the same mathematical principles underpinning the aforementioned finding, also imply an inherent capability of AFDM to reap the full uncoded diversity of static linear time-invariant (LTI) channels, demonstrating that this simple upgrade taps into previously undiscovered strengths of multicarrier waveforms.

This letter studies distributed stochastic optimization over a peer-to-peer network when agents can query only zeroth-order function values. We propose ZOOM-PB, a coordinate-sampling distributed zeroth-order method equipped with a fractional-power powerball map. Unlike existing distributed zeroth-order methods that mainly refine gradient estimation or introduce primal--dual tracking, the proposed mechanism acts as a nonlinear feedback gain on the estimated gradient: it amplifies weak signals in flat regions and attenuates large stochastic estimates without adding transmitted states. Under standard smoothness, oracle-variance, and network-connectivity assumptions, ZOOM-PB achieves the leading nonconvex stationarity rate $\mathcal{O}(\sqrt{p/(nT)})$, where $p$ is the decision dimension, $n$ is the number of agents, and $T$ is the iteration horizon. Under the Polyak--Łojasiewicz condition, it further attains the leading objective residual rate $\mathcal{O}(p/(nT))$. Thus the method preserves the known distributed ZO order while changing the finite-time behavior through a local nonlinear control gain. Simulations on black-box learning and sensor-driven UAV source seeking show faster empirical convergence in weak-signal regimes.

Nowadays, the convergence of mobile edge computing (MEC) and vehicular networks has emerged as a vital enabler for the ever-increasing intelligent onboard applications. This paper proposes a multi-tier task offloading mechanism for MEC-enabled vehicular networks leveraging vehicle-to-everything (V2X) communications. The study focuses on applications with sequential subtasks and explores the collaboration of two tiers. In the Vehicle Tier, the requesting vehicle (RV)-service vehicle (SV) matching scheme and the inter-vehicle collaborative computation are studied, with joint optimization of task offloading decision, communication, and computing resource allocation to minimize energy consumption while satisfying delay requirements. In the Roadside Unit (RSU) Tier, collaboration among RSUs is investigated to further address multi-access issues of uplink subchannels and computing resources for serving unmatched RVs. To tackle this intricate problem, a layered optimization framework is first proposed to obtain task offloading decisions and optimal continuous resource allocation, after which a subchannel allocation scheme is designed to recover the discrete solution with low complexity. Extensive experiments are conducted to demonstrate that the proposed method reduces average energy consumption by at least 15% compared with recent utility maximization and energy cost minimization benchmarks under varying task delay requirements and vehicle scales.

Computational Music Generation is evolving towards non-conventional styles, demanding methods that enable precise and controllable blending of diverse music elements. In this work, we present a method for fine grained control using inference-time interventions on an autoregressive generative transformer, MusicGen. Through our approach, we achieve genre control by steering the residual stream using weights of a linear probe on it. By framing activation steering as a human-controllable interaction, our work highlights how interpretable model behaviors can empower in co-creative music this http URL samples demonstrating our method are available on our demo page.

Wireless communication is evolving into an agent era, where large-scale agents with inherent embodied intelligence are not just users but active participants. The perfect combination of wireless communication and embodied intelligence can achieve a synergetic empowerment and greatly facilitate the development of agent communication. An overview of this synergetic empowerment is presented, framing it as a co-evolutionary process that transforms wireless communication from a simple utility into the digital nervous system of a collective intelligence, while simultaneously elevating isolated agents into a unified superorganism with emergent capabilities far exceeding individual contributions. Moreover, we elaborate how embodied intelligence and wireless communication mutually benefit each other through the lens of the perception-cognition-execution (PCE) loop, revealing a fundamental duality where each PCE stage both challenges network capacity and creates unprecedented opportunities for system-wide optimization. Furthermore, critical open issues and future research directions are identified.

We present the Social Influence Game (SIG), a framework for modeling adversarial persuasion in social networks with an arbitrary number of competing players. Our goal is to provide a tractable and interpretable model of contested influence that scales to large systems while capturing the structural leverage points of networks. Each player allocates influence from a fixed budget to steer opinions that evolve under DeGroot dynamics, and we prove that the resulting optimization problem is a difference-of-convex program. To enable scalability, we develop an Iterated Linear (IL) solver that approximates player objectives with linear programs. In experiments on random and archetypical networks, IL achieves solutions within 7% of nonlinear solvers while being over 10x faster, scaling to large social networks. This paper lays a foundation for asymptotic analysis of contested influence in complex networks.

General-purpose object detectors face fundamental structural limitations when applied to ship detection in satellite imagery, where the ship scale distribution is concentrated at small sizes and high aspect ratios. In conventional You Only Look Once architectures, the deepest feature pyramid level (stride 32) compresses narrow vessels into sub-pixel representations, causing severe spatial feature dilution and compromising accurate ship boundary regression. We propose Less is More YOLO, a streamlined detector built upon the extra-large variant of YOLOv9, to address these domain-specific structural conflicts. From a statistical analysis of ship scale distributions across four major benchmarks (SODA-A, DOTA-v1.5, FAIR1M-v2.0, and ShipRSImageNet), we introduce a Pyramid Level Shift Strategy that shifts the detection head from strides 8, 16, and 32 to strides 4, 8, and 16. This shift satisfies a spatial representability condition derived from the Nyquist-Shannon principle for the narrowest targets, while eliminating the computational redundancy of the deepest pyramid level. To further stabilize training on high-resolution satellite inputs, we incorporate a group-normalized auxiliary projection module that introduces Group Normalization into the projection path, mitigating gradient instability in memory-constrained micro-batch regimes. Validated on these four datasets, our detector attains an mAP_{50-95} of 0.600 with only 21.16 million parameters, a 64.1% reduction from the extra-large YOLOv9 baseline (58.99 million). Despite this compact size, our model surpasses state-of-the-art detectors up to three times larger, validating that a well-targeted pyramid level shift achieves a "Less is More" balance between accuracy and efficiency. The code is available at this https URL.

Learning the structure of directed acyclic graphs (DAGs) from observational data is a central problem in causal discovery, statistical signal processing, and machine learning. Under a linear Gaussian structural equation model (SEM) with equal noise variances, the problem is identifiable and we show that the ensemble precision matrix of the observations exhibits a distinctive structure that facilitates DAG recovery. Exploiting this property, we propose BUILD (Bottom-Up Inference of Linear DAGs), a deterministic stepwise algorithm that identifies leaf nodes and their parents, then prunes the leaves by removing incident edges to proceed to the next step, exactly reconstructing the DAG from the true precision matrix. In practice, precision matrices must be estimated from finite data, and ill-conditioning may lead to error accumulation across BUILD steps. As a mitigation strategy, we periodically re-estimate the precision matrix (with less variables as leaves are pruned), trading off runtime for enhanced robustness. Reproducible results on challenging synthetic benchmarks demonstrate that BUILD compares favorably to state-of-the-art DAG learning algorithms, while offering an explicit handle on complexity.

Speech tokenizers are a key building block of fully discrete Speech LLMs. Existing tokenizers either prioritize semantic encoding, fuse semantic content with acoustic style inseparably, or achieve incomplete semantic-acoustic disentanglement. To achieve better disentanglement, we propose \textbf{DSA-Tokenizer}, which explicitly disentangles speech into discrete semantic and acoustic tokens via distinct optimization constraints. Specifically, semantic tokens are supervised by ASR to capture linguistic content, while acoustic tokens focus on mel-spectrograms restoration to encode style. We further introduce a hierarchical Flow Matching decoder and a joint reconstruction-context inpainting training strategy, allowing the model to support both high-fidelity reconstruction and cross-utterance voice clone. To speed up inference, we distill the DiT decoder to reduce sampling steps of inference to 4 and improve synthesis quality with GAN fine-tuning. Experiments demonstrate that DSA-Tokenizer provides strong semantic-acoustic disentanglement, reliable controllable voice cloning, and efficient high-fidelity generation with low WER/CER. Moreover, our results suggest that disentangled tokenization provides a more effective interface for downstream large-model speech generation. Audio samples are avaialble at this https URL.

Time-optimal control for triple integrator under full box constraints is a fundamental problem in the field of optimal control, which has been widely applied in the industry. However, scenarios involving asymmetric constraints, non-stationary boundary conditions, and active position constraints pose significant challenges. This paper provides a complete characterization of time-optimal switching surfaces for the problem, leading to novel insights into the geometric structure of the optimal control. The active condition of position constraints is derived, which is absent from the literature. An efficient algorithm is proposed, capable of planning time-optimal trajectories under asymmetric full constraints and arbitrary boundary states, with a 100% success rate. Computational time for each trajectory is within approximately 10$\mu$s, achieving a 5-order-of-magnitude reduction compared to optimization-based baselines.

We present a reproducible benchmark for evaluating sim-to-real transfer of Multi-Agent Reinforcement Learning (MARL) policies for Connected and Automated Vehicles (CAVs). The platform, based on the Cyber-Physical Mobility Lab (CPM Lab) [1], integrates simulation, a high-fidelity digital twin, and a physical testbed, enabling structured zero-shot evaluation of MARL motion-planning policies. We demonstrate its use by deploying a SigmaRL-trained policy [2] across all three domains, revealing two complementary sources of performance degradation: architectural differences between simulation and hardware control stacks, and the sim-to-real gap induced by increasing environmental realism. The open-source setup enables systematic analysis of sim-to-real challenges in MARL under realistic, reproducible conditions.

Many robotic systems allow independent control of position and orientation (pose), including omnidirectional aerial vehicles, underwater robots, and manipulator end-effectors. In many applications, these systems must follow a continuous sequence of poses, leading to either trajectory-tracking or path following formulations. Compared to trajectory-tracking, path following offers important practical advantages. In particular, we focus on the problem of path following on Lie groups. Considering the robots as rigid bodies moving in the 3D space, this path-following problem can be posed as a problem of designing guiding vector fields on the matrix Lie group SE(3). In this paper, we develop a general vector-field framework for path following on connected matrix Lie groups, of which SE(3) is a prominent special case. The proposed vector field guarantees convergence to a desired parametric curve from almost all initial conditions while ensuring continuous motion along the path. Furthermore, another interesting feature is that, as opposed to previous works, the control input is "minimal" in terms of representation and closer to the engineering application (e.g., the body twist in the case SE(3)). After establishing the general case, the framework is then specialized to SE(3), of special interest in robotics, yielding an efficient algorithm suitable for real-time robotic control. Experiments with a robotic manipulator tracking complex pose paths demonstrate the effectiveness of the approach. An open-source implementation is also provided.

High intermittent renewable penetration in the energy mix presents challenges in robustness for the management of power systems' operation. If a tail realization of the distribution of weather yields a prolonged period of time during which solar irradiation and wind speed are insufficient for satisfying energy demand, then it becomes critical to ramp up the generation of conventional power plants with adequate foresight. This event trigger is costly, and inaccurate forecasting can either be wasteful or yield catastrophic undersupply. This encourages particular attention to accurate modeling of the noise and the resulting dynamics within the aforementioned scenario. In this work we present a method for rare event-aware control of power systems using multi-stage scenario-based stochastic programming. A Fleming-Viot particle approach is used to bias the scenario generation towards rare realizations of very low wind power, in order to obtain a cost-effective control of conventional power plants that is robust under prolonged renewable energy shortfalls.

This letter presents an optimal-transport (OT)-driven, distributionally robust attack detection algorithm, OT-DETECT, for cyber-physical systems (CPS) modeled as partially observed linear stochastic systems. The underlying detection problem is formulated as a minmax optimization problem using 1-Wasserstein ambiguity sets constructed from observer residuals under both the nominal (attack-free) and attacked regimes, and show that the minmax detection problem can be reduced to a finite-dimensional linear program for computing the worst-case distribution (WCD). Off-support residuals are handled via a kernel-smoothed score function that drives a CUSUM procedure for sequential detection. We also establish a non-asymptotic tail bound on the false-positive error of the CUSUM statistic under the nominal (attack-free) condition, under mild assumptions. Numerical illustrations are provided to evaluate the robustness properties of OT-DETECT.

We consider nonlinear model predictive control (MPC) schemes without stabilizing terminal conditions, where the model used in the optimization step is generated based on input-output data only. We establish exponential stability for sufficiently long prediction horizons assuming exponential stabilizability and a proportional error bound. Moreover, we verify the imposed condition on the approximation using kernel interpolation and demonstrate the practical applicability to nonlinear systems by numerical simulations.

The Prisoner's Dilemma, zero-sum games, LQR team problems, and differential games have shaped game theory in controls for decades, but the field's most pressing adversarial challenges demand a richer framework, and its name is Colonel Blotto. Strategic adversarial constraints represent a fundamental consideration in control systems, from cybersecurity defense to infrastructure protection. Colonel Blotto games, despite their direct relevance to such applications, remain underutilized in the controls community relative to other game-theoretic approaches. This article aims to close that gap for the controls community. Indeed, theoretical advances within the last two decades have spurred a resurgence of interest and enabled their applications across several domains. In this article, we introduce the Colonel Blotto framework, survey key analytical and computational results, and demonstrate how problems spanning cybersecurity, network defense, and multi-agent systems fit naturally within this structure. Three research directions are examined in depth: interdependent contest objectives that capture networked vulnerabilities, alternate winning rules that model partial rewards and structural asymmetries, and multi-agent competitive environments involving coalition formation and strategic concessions. Taken together, these directions reveal a framework that is both practically deployable and rich enough to capture the strategic complexity inherent in adversarial resource allocation.

Stem retrieval, the task of matching missing stems to a given audio submix, is a key challenge currently limited by models that discard temporal information. We introduce PHALAR, a contrastive framework achieving a relative accuracy increase of up to $\approx 70\%$ over the state-of-the-art while requiring $<50\%$ of the parameters and a 7$\times$ training speedup. By utilizing a Learned Spectral Pooling layer and a complex-valued head, PHALAR enforces pitch-equivariant and phase-equivariant biases. PHALAR establishes new retrieval state-of-the-art across MoisesDB, Slakh, and ChocoChorales, correlating significantly higher with human coherence judgment than semantic baselines. Finally, zero-shot beat tracking and linear chord probing confirm that PHALAR captures robust musical structures beyond the retrieval task.

This paper studies autonomous generative AI agents in multi-echelon supply chains using the MIT Beer Game. We identify four inference-time levers that shape performance: model selection, policies and guardrails, centralized data sharing, and prompt engineering. Model capability is the dominant factor: an out-of-the-box reasoning model exceeds human-level performance, and optimized reasoning models reduce costs by up to 67% relative to human teams. However, strong average performance masks substantial reliability risks. We introduce agent bullwhip: the amplification of run-to-run decision instability in autonomous multi-echelon systems. A central component is decision bullwhip, the portion of order variability generated by stochastic agent decisions rather than by changes in customer demand. We show that decision instability can amplify both across facilities at a fixed point in time and within the same facility over time, even when the demand path is held fixed. Repeated sampling, a natural test-time remedy, fails to meaningfully reduce this instability, suggesting that reliability requires changing the underlying decision policy rather than merely averaging over model outputs. To address this limitation, we propose a Group Relative Policy Optimization (GRPO)-based reinforcement-learning post-training framework that trains a shared base LLM using system-level supply-chain rewards. Post-training substantially reduces tail events, curtails agent bullwhip, and improves the reliability of autonomous supply-chain agents.