Papers

Background: Diagnostic PET image quality depends on the administered activity and acquisition time. However, minimizing these variables is desirable to reduce patient radiation exposure and radiopharmaceutical costs. PETfectior is an artificial intelligence-based software that processes PET scans and increases signal-to-noise ratio, obtaining high-quality images from low-count-rate images. We perform an initial clinical validation of PETfectior on images acquired with half of the counting statistics required to meet the most recent EANM quantitative standards for 18F-FDG PET, evaluating lesions detectability, quantitative performance and image quality. Materials and methods: 258 patients referred for 18F-FDG PET/CT were prospectively included. The standard-of-care scans (100% scans) were acquired and reconstructed according to EARL standards 2. Half-counting-statistics versions were generated from list-mode data and processed with PETfecftior (50%+PETfectior scans). All oncologic lesions were segmented on both PET/CT versions, manually or automatically, and lesions detectability was evaluated. The SUVmax of the lesions was measured and the quantitative concordance of 50%+PETfectior and 100% images was evaluated. Subjective image quality was visually assessed by two experienced physicians. Results: 1649 lesions were detected in a total of 198 studies. The 50%+PETfectior images showed high sensitivity for lesion detection (99.9%) and only 1 false positive was detected. The SUVmax measured in 100% and 50%+PETfectior images agreed within 12.5% (95% limits of agreement), with a bias of -1.01%. Image quality of the 50%+PETfectior images was rated equal to or better than the standard-of-care images. Conclusion: PETfectior can safely be used in clinical practice at half counting statistics, with high sensitivity and specificity, low quantitative bias and high subjective image quality.

Reachability analysis has been a prominent way to provide safety guarantees for neurally controlled autonomous systems, but its direct application to neural perception components is infeasible due to imperfect or intractable perception models. Typically, this issue has been bypassed by complementing reachability with statistical analysis of perception error, say with conformal prediction (CP). However, existing CP methods for time-series data often provide conservative bounds. The corresponding error accumulation over time has made it challenging to combine statistical bounds with symbolic reachability in a way that is provable, scalable, and minimally conservative. To reduce conservatism and improve scalability, our key insight is that perception error varies significantly with the system's dynamical state. This article proposes state-dependent conformal prediction, which exploits that dependency in constructing tight high-confidence bounds on perception error. Based on this idea, we provide an approach to partition the state space, using a genetic algorithm, so as to optimize the tightness of conformal bounds. Finally, since using these bounds in reachability analysis leads to additional uncertainty and branching in the resulting hybrid system, we propose a branch-merging reachability algorithm that trades off uncertainty for scalability so as to enable scalable and tight verification. The evaluation of our verification methodology on two complementary case studies demonstrates reduced conservatism compared to the state of the art.

Virtual acoustic environments enable the creation and simulation of realistic and eco-logically valid daily-life situations vital for hearing research and audiology. Reverberant indoor environments are particularly important. For real-time applications, room acous-tics simulation requires simplifications, however, the necessary acoustic level of detail (ALOD) remains unclear in order to capture all perceptually relevant effects. This study examines the impact of varying ALOD in simulations of three real environments: a living room with a coupled kitchen, a pub, and an underground station. ALOD was varied by generating different numbers of image sources for early reflections, or by excluding geo-metrical room details specific for each environment. Simulations were perceptually eval-uated using headphones in comparison to binaural room impulse responses measured with a dummy head in the corresponding real environments, or by using loudspeakers. The study assessed the perceived overall difference for a pulse stimulus, a played electric bass and a speech token. Additionally, plausibility, speech intelligibility, and externaliza-tion were evaluated. Results indicate that a strong reduction in ALOD is feasible while maintaining similar plausibility, speech intelligibility, and externalization as with dummy head recordings. The number and accuracy of early reflections appear less relevant, pro-vided diffuse late reverberation is appropriately represented.

This letter proposes an AC/DC frequency-dependent power flow Jacobian analysis to identify the system support capabilities. In addition, the analyses reveal that system support capabilities do not necessarily enhance the system stability margin, suggesting that technical requirements of narrow-frequency-band and AC-side focused specifications may not lead to the expected performance of GFM.

Stochastic Nonlinear Optimal Control (SNOC) seeks to minimize a cost function that accounts for random disturbances acting on a nonlinear dynamical system. Since the expectation over all disturbances is generally intractable, a common surrogate is the empirical cost, obtained by averaging over a finite dataset of sampled noise realizations. This substitution, however, introduces the challenge of guaranteeing performance under unseen disturbances. The issue is particularly severe when the dataset is limited, as the trained controllers may overfit, leading to substantial gaps between their empirical cost and the deployment cost. In this work, we develop a PAC-Bayesian framework that establishes rigorous generalization bounds for SNOC. Building on these bounds, we propose a principled controller design method that balances empirical performance and prior knowledge. To ensure tractability, we derive computationally efficient relaxations of the bounds and employ approximate inference methods. Our framework further leverages expressive neural controller parameterizations, guaranteeing closed-loop stability. Through simulated examples, we highlight how prior knowledge can be incorporated into control design and how more reliable controllers can be synthesized for cooperative robotics.

Model-based filtering is often carried out while subject to an imperfect model, as learning partially-observable stochastic systems remains a challenge. Recent work on Bayesian inference found that tempering the likelihood or full posterior of an imperfect model can improve predictive accuracy, as measured by expected negative log likelihood. In this paper, we develop the tempered Bayes filter, improving estimation performance through both of the aforementioned, and one newly introduced, modalities. The result admits a recursive implementation with a computational complexity no higher than that of the original Bayes filter. Our analysis reveals that – besides the well-known fact in the field of Bayesian inference that likelihood tempering affects the balance between prior and likelihood – full-posterior tempering tunes the level of entropy in the final belief distribution. We further find that a region of the tempering space can be understood as interpolating between the Bayes- and MAP filters, recovering these as special cases. Analytical results further establish conditions under which a tempered Bayes filter achieves improved predictive performance. Specializing the results to the linear Gaussian case, we obtain the tempered Kalman filter. In this context, we interpret how the parameters affect the Kalman state estimate and covariance propagation. Empirical results confirm that our method consistently improves predictive accuracy over the Bayes filter baseline.

This paper presents the modeling of autonomous vehicles with high maneuverability used in an experimental framework for educational purposes. Since standard bicycle models typically neglect wide steering angles, we develop modified planar bicycle models and combine them with both parametric and non-parametric identification techniques that progressively incorporate physical knowledge. The resulting models are systematically compared to evaluate the tradeoff between model accuracy and computational requirements, showing that physics-informed neural network models surpass the purely physical baseline in accuracy at lower computational cost.

This paper presents a Gain-Scheduled Data-Enabled Predictive Control (GS-DeePC) framework for nonlinear systems based on multiple locally linear data representations. Instead of relying on a single global Hankel matrix, the operating range of a measurable scheduling variable is partitioned into regions, and regional Hankel matrices are constructed from persistently exciting data. To ensure smooth transitions between linearization regions and suppress region-induced chattering, composite regions are introduced, merging neighboring data sets and enabling a robust switching mechanism. The proposed method maintains the original DeePC problem structure and can achieve reduced computational complexity by requiring only short, locally informative data sequences. Extensive experiments on a nonlinear DC-motor with an unbalanced disc demonstrate the significantly improved control performance compared to standard DeePC.

Synthetic aperture radar (SAR) imaging plays a critical role in all-weather, day-and-night remote sensing, yet reconstruction is often challenged by noise, undersampling, and complex scattering scenarios. Conventional methods, including matched filtering and sparsity-based compressed sensing, are limited in capturing intricate scene structures and frequently suffer from artifacts, elevated sidelobes, and loss of fine details. Recent diffusion models have demonstrated superior capability in representing high-order priors; however, existing diffusion-based SAR methods still yield degraded reconstructions due to oversimplified likelihood approximations in guided sampling. In this work, we propose a diffusion-driven split Gibbs sampling framework for SAR reconstruction, rigorously integrating measurement fidelity with learned diffusion priors. By alternately performing likelihood- and prior-driven updates via proximal sampling, this method ensures progressive convergence toward the true posterior while fully leveraging the expressive power of diffusion priors. Extensive experiments on simulated and Sentinel-1A datasets demonstrate substantial performance improvements: over 7 dB average PSNR gain in simulations, along with significant sidelobe suppression (MPLSR +2.96 dB, MISLR +11.5 dB) with respect to the best baseline result. On real-world Sentinel-1A data, the method achieves an average PSNR gain of 1.6 dB while effectively reducing artifacts and preserving scene details, including ridges, edges, and fine textures. These results underscore the potential of the adapted framework as a robust and generalizable solution for high-fidelity SAR imaging across diverse sensing scenarios.

Accelerated brain aging has been consistently reported in patients with schizophrenia. Over the past decade, these findings have been replicated using the Brain Age paradigm, which applies machine learning techniques to estimate brain age from neuroimaging data. This approach yields a single index, the Brain Age Gap, defined as the difference between predicted and chronological age. Nevertheless, both the progressive nature of this phenomenon and the potential role of antipsychotic medication remain unclear. To investigate its progression, we compared the Brain Age Gap between individuals experiencing a first episode of psychosis and healthy controls using ANCOVA, adjusting for age, sex, body mass index, and estimated total intracranial volume. To enhance the robustness of our findings, we employed two distinct models: a transformer-inspired model based on harmonized volumetric brain features extracted with FastSurfer, and a previously trained deep learning model. To assess the potential effect of medication, we further compared bipolar patients who received antipsychotic treatment with those who did not. Mann-Whitney U test consistently showed that medicated bipolar patients did not exhibit a significantly larger Brain Age Gap. Both models converge on the conclusion that accelerated brain aging is unlikely to be explained by antipsychotic medication alone. Longitudinal studies are therefore required to clarify the temporal dynamics of brain aging in schizophrenia.

This paper describes the UZH-CL system submitted to the FAME2026 Challenge. The challenge focuses on cross-modal verification under unique multilingual conditions, specifically unseen and unheard languages. Our approach investigates two distinct architectures, consisting of a baseline dual-encoder system trained from scratch using contrastive and orthogonal projection losses, and a foundation model approach leveraging ImageBind with LoRA. To address the data scarcity and language constraints of the challenge, we curated an external Arabic dataset from VoxBlink. Our best-performing system, ImageBind-LoRA, demonstrates remarkable cross-lingual generalization: despite being fine-tuned exclusively on Arabic audio, it achieved an EER of 24.73% on the evaluation set (English and German), securing 2nd place in the competition.

The ability to construct channel knowledge map (CKM) with high precision is essential for environment awareness in 6G wireless systems. However, most existing CKM construction methods formulate the task as an image super-resolution or generation problem, thereby employing models originally developed for computer vision. As a result, the generated CKMs often fail to capture the underlying physical characteristics of wireless propagation. In this paper, considering that acquiring channel observations incurs non-negligible time and cost, we focus on constructing CKM for large-scale fading scenarios without relying on prior observations, and we design three physics-based constraints to characterize the spatial distribution patterns of large-scale fading. By integrating these physical constraints with state-of-the-art diffusion model that possesses superior generative capability, a physics-inspired diffusion model for CKM construction is proposed. Following this motivation, we derive the loss function of the diffusion model augmented with physics-based constraint terms and further design the training and generation framework for the proposed physics-inspired CKM generation diffusion model. Extensive experiments show that our approach outperforms all existing methods in terms of construction accuracy. Moreover, the proposed model provides a unified and effective framework with strong potential for generating diverse, accurate, and physically consistent CKM.

Physics-Informed Neural Networks (PINNs) have advanced the data-driven solution of differential equations (DEs) in dynamic physical systems, yet challenges remain in explainability, scalability, and architectural complexity. This paper presents a Generalizable Physics-Informed Fourier Neural Network (G-PIFNN) framework that enhances PINN architectures for efficient and interpretable electrical circuit analysis. The proposed G-PIFNN introduces three key advancements: (1) improved performance and interpretability via a physics activation function (PAF) and a lightweight Physics-Informed Fourier Neural Network (PIFNN) architecture; (2) automated, bond graph (BG) based formulation of physics-informed loss functions for systematic differential equation generation; and (3) integration of intra-circuit and cross-circuit class transfer learning (TL) strategies, enabling unsupervised fine-tuning for rapid adaptation to varying circuit topologies. Numerical simulations demonstrate that G-PIFNN achieves significantly better predictive performance and generalization across diverse circuit classes, while significantly reducing the number of trainable parameters compared to standard PINNs.

In Real-Time, Feedback-Driven Single Particle Tracking methods, measurements of the emission intensity from a labeled, nanometer-scale particle are used in a feedback loop to track the motion of the particle as it moves inside its native environment, including within living cells. In this work, we take advantage of Point Spread Function (PSF) engineering techniques that encode the axial position of the particle into the shape of the PSF in the focal plane to eliminate the need for out-of-focal-plane measurements, reducing the complexity of implementation and decreasing the overall measurement time of the control loop. Specifically, we used the Double Helix PSF (DH-PSF) in which a single fluorescent source gives rise to two lobes in the image plane with the lobes rotating in the plane as the particle moves along the optical axis. We designed simple estimators of the relative error between the particle and the tracker, and a simple proportional feedback controller to regulate that error to zero. We explored the efficacy of the approach through simulation studies, demonstrating the tracking of fast-moving particles (with diffusion coefficients up to 10 {\mu\text{m}^2/\text{s}}) over long time periods (multiple seconds).

Mountain huts are buildings located at high altitude, providing shelter and a place for hikers. Energy supply on mountain huts remains an open issue. Using renewable energies could be an appropriate solution. Tianchi Lodge, a famous mountain hut in Taiwan, has operated an off-grid solar energy storage system with lithium iron phosphate (LFP) batteries since 2020. In this case report, the energy architecture, detailed descriptions, and historical status of the system are provided.

This paper employs modal analysis to study the core inertial dynamics of governor-controlled synchronous generators (GC-SG), droop-based grid-forming (GFM) converters, and their most fundamental interactions. The results indicate that even in the simplest cases, the prevailing industry paradigm of emulating legacy GC-SG behaviour in GFM converters (high inertia to slow down the system and large droop to increase damping) could be a suboptimal policy. It is shown that GC-SGs exhibit a fundamental trade-off: adequate damping of the turbine-governor mode requires large droop constants, inevitably increasing steady-state frequency deviation and dependence on secondary regulation. In contrast, droop-based GFM converters invert this relationship: decreasing the droop constant simultaneously reduces steady-state frequency deviations and increases damping, while allowing virtual inertia to be freely chosen. When two GC-SGs are coupled, the poorly damped electromechanical swing mode emerges. Results show that replacing one GC-SG with a GFM converter of equivalent droop and inertia already significantly improves damping of both swing and turbine-governor modes. Counter-intuitively, further and remarkable damping gains are achieved by substantially lowering the GFM virtual inertia constant. These findings suggest that current industry trends may be constraining the potential benefits of Inverter Based Resources (IBRs). Optimal stability and performance are instead obtained with low droop and low virtual inertia, yielding tightly bounded frequency variations and strongly-damped electromechanical modes. The results indicate a need to re-evaluate GFM control design principles and emerging grid-code requirements.

This work introduces a new task, text-conditioned selective video-to-audio (V2A) generation, which produces only the user-intended sound from a multi-object video. This capability is especially crucial in multimedia production, where audio tracks are handled individually for each sound source for precise editing, mixing, and creative control. We propose SELVA, a novel text-conditioned V2A model that treats the text prompt as an explicit selector to distinctly extract prompt-relevant sound-source visual features from the video encoder. To suppress text-irrelevant activations with efficient video encoder finetuning, the proposed supplementary tokens promote cross-attention to yield robust semantic and temporal grounding. SELVA further employs an autonomous video-mixing scheme in a self-supervised manner to overcome the lack of mono audio track supervision. We evaluate SELVA on VGG-MONOAUDIO, a curated benchmark of clean single-source videos for such a task. Extensive experiments and ablations consistently verify its effectiveness across audio quality, semantic alignment, and temporal synchronization.

Massive multiple-input multiple-output (MIMO) offers substantial spectral-efficiency gains, but scaling to very large antenna arrays with conventional all-digital and hybrid beamforming architectures quickly results in excessively high costs and power consumption. Low-cost, switch-based architectures have recently emerged as a potential alternative. However, prior studies rely on simplified models that ignore (among others) antenna coupling, radiation patterns, and matching losses, resulting in inaccurate performance predictions. In this paper, we use a physically consistent electromagnetic modeling framework to analyze an ultra-dense patch-antenna array architecture that performs joint beamforming and matching using networks of inexpensive RF switches. Our results demonstrate that simple, switch-based beamforming architectures can approach the antenna-gain of all-digital solutions at significantly lower cost and complexity.

Transfer learning aims to optimize performance in a target task by learning from a related source problem. In this work, we propose an efficient transfer learning method using a tensor kernel machine. Our method takes inspiration from the adaptive SVM and hence transfers 'knowledge' from the source to the 'adapted' model via regularization. The main advantage of using tensor kernel machines is that they leverage low-rank tensor networks to learn a compact non-linear model in the primal domain. This allows for a more efficient adaptation without adding more parameters to the model. To demonstrate the effectiveness of our approach, we apply the adaptive tensor kernel machine (Adapt-TKM) to seizure detection on behind-the-ear EEG. By personalizing patient-independent models with a small amount of patient-specific data, the patient-adapted model (which utilizes the Adapt-TKM), achieves better performance compared to the patient-independent and fully patient-specific models. Notably, it is able to do so while requiring around 100 times fewer parameters than the adaptive SVM model, leading to a correspondingly faster inference speed. This makes the Adapt-TKM especially useful for resource-constrained wearable devices.

Spoken conversational agents are converging toward voice-native LLMs. This tutorial distills the path from cascaded ASR/NLU to end-to-end, retrieval-and vision-grounded systems. We frame adaptation of text LLMs to audio, cross-modal alignment, and joint speech-text training; review datasets, metrics, and robustness across accents and compare design choices (cascaded vs. E2E, post-ASR correction, streaming). We link industrial assistants to current open-domain and task-oriented agents, highlight reproducible baselines, and outline open problems in privacy, safety, and evaluation. Attendees leave with practical recipes and a clear systems-level roadmap.

This article deals with the implementation of the Smith Predictor for state feedback control in state space representation. The desired control law, obtained using partial differential equations and backstepping control, contains an integral term that has to be approximated for implementation. In this article, we propose a new way to implement this control law using a dynamic controller. The control law is composed of a state feedback term and a dynamic term that approaches the integral term that has to be estimated for implementation. Using a Lyapunov functional, we provide sufficient conditions, in terms of a linear matrix inequality, to guarantee that the closed-loop system is stable when the proposed control law is applied. We use three examples, taken from the literature, to show the benefits of the proposed approach.

In multi-user multiple-input multiple-output (MU-MIMO) systems, the non-linear behavior of the power amplifiers (PAs) may cause degradation of the linear precoding schemes dealing with interference between user equipments (UEs), e.g., the zero-forcing (ZF) precoder. One way to minimize this effect is to use digital-pre-distortion (DPD) modules to linearize the PAs. However, using perfect DPD modules is costly and it may incur significant power consumption. As an alternative, we consider the problem of characterizing non-linearity-aware ZF (NLA-ZF) precoding schemes, hereby defined as linear precoders that achieve perfect interference cancellation in the presence of PA non-linearity by exploiting knowledge of this non-linear response. We provide initial iterative solutions that allow achieving NLA-ZF (up to adjustable tolerance) in a two-UE downlink MU-MIMO scenario where the base station (BS) has an even number of antennas, and each antenna is connected to a PA exhibiting third-order memory-less non-linear behavior. The proposed approach allows for performance gains in scenarios with significant residual interference.

Accurate beam prediction is essential for maintaining reliable links and high spectral efficiency in dynamic low-altitude wireless networks. However, existing approaches often fail to capture the deep correlations across heterogeneous sensing modalities, limiting their adaptability in complex three-dimensional environments. To overcome these challenges, we propose a multi-modal predictive beamforming method based on a cross-attention fusion mechanism that jointly leverages visual and structured sensor data. The proposed model utilizes a Convolutional Neural Network (CNN) to learn multi-scale spatial feature hierarchies from visual images and a Transformer encoder to capture cross-dimensional dependencies within sensor data. Then, a cross-attention fusion module is introduced to integrate complementary information between the two modalities, generating a unified and discriminative representation for accurate beam prediction. Through experimental evaluations conducted on a real-world dataset, our method reaches 79.7% Top-1 accuracy and 99.3% Top-3 accuracy, surpassing the 3D ResNet-Transformer baseline by 4.4%-23.2% across Top-1 to Top-5 metrics. These results verify that multi-modal cross-attention fusion is effective for intelligent beam selection in dynamic low-altitude wireless networks.

Intersections present significant challenges in traffic management, where ensuring safety and efficiency is essential for effective flow. However, these goals are often achieved at the expense of fairness, which is critical for trustworthiness and long-term sustainability. This paper investigates how the timing of centralized intervention affects the management of autonomous agents at a signal-less, orthogonal intersection, while satisfying safety constraints, evaluating efficiency, and ensuring fairness. A mixed-integer linear programming (MILP) approach is used to optimize agent coordination within a circular control zone centered at the intersection. We introduce the concept of fairness, measured via pairwise reversal counts, and incorporate fairness constraints into the MILP framework. We then study the relationship between fairness and system efficiency and its impact on platoon formation. Finally, simulation studies analyze the effectiveness of early versus late intervention strategies and fairness-aware control, focusing on safe, efficient, and robust management of agents within the control zone.

In next-generation wireless communication systems, the newly designated upper mid-band has attracted considerable attention, also called frequency range 3 (FR3), highlighting the need for downlink (DL) transmission design, which fundamentally relies on accurate CSI. However, CSI acquisition in FR3 systems faces significant challenges: the increased number of antennas and wider transmission bandwidth introduces prohibitive training overhead with traditional estimation approaches, as each probing captures only incomplete spatial-frequency observation, while higher carrier frequencies lead to faster temporal channel variation. To address these challenges, we propose a novel CSI acquisition framework that integrates CSI feedback, uplink (UL) and DL channel estimation, as well as channel prediction in the FR3 TDD massive MIMO systems. Specifically, we first develop the Joint UL and DL Channel Estimation Network (JUDCEN) to fuse incomplete observations based on the SRSs and CSI-RSs. By exploiting the complementary characteristics of preliminary UL and DL estimation features, obtained through initial UL estimation and quantized-feedback-assisted DL estimation, it enables full CSI reconstruction in the spatial domain. To mitigate the performance degradation in the feedback process, we propose the Transformer-MLP CSI Feedback Network (TMCFN), employing an MLP-based module to jointly exploit angle- and delay-domain features. Building upon the reconstructed full CSI, we further develop the Mamba-based Channel Prediction Network (MCPN), which exploits selective state-space model (SSM) mechanism to capture long-range temporal dynamics in the angle-delay domain for future CSI prediction. Simulation results demonstrate that the proposed framework consistently outperforms benchmarks in both CSI acquisition accuracy and transmission spectral efficiency with lower computational complexity.

Cell-free massive MIMO (CF-mMIMO) has emerged as a promising technology for next generation wireless systems, combining the benefits of distributed antenna systems (DAS) and traditional MIMO technology. In this work, we present the first extensive channel measurements for CF-mMIMO in the mid-band (FR3, 6-24 GHz), using a virtual widely distributed antenna array comprising 512 elements in the urban Macrocell (UMa) environment. Based on the measurement data, this paper compares the channel capacity of CF-mMIMO and Conventional mMIMO under both line-of-sight (LOS) and non-line-of-sight (NLOS) conditions across a range of signal-to-noise ratios (SNRs). We then analyze how channel capacity varies with Rx positions from the perspectives of the full array and of individual subarrays. Finally, we conclude that the 64-element array configuration yields the greatest advantage in channel capacity for CF-mMIMO in the measurement environment considered, with gains of 14.02% under LOS and 24.61% under NLOS conditions. This in-depth analysis of channel capacity in the FR3 band provides critical insights for optimizing CF-mMIMO systems in next generation wireless networks.

This letter addresses the joint air corridor planning and base station (BS) deployment problem for low-altitude integrated sensing and communication (ISAC) networks. In the considered system, unmanned aerial vehicles (UAVs) operate within a structured air corridor composed of connected cubic segments, and multiple BSs need to be selectively deployed at a set of candidate locations to ensure both sensing and communication coverage throughout the corridor. In particular, we leverage the channel knowledge map (CKM) to characterize wireless channels for candidate BS sites prior to deployment, thereby facilitating the offline planning. Under this setup, we minimize the system cost in terms of the weighted sum of the air corridor length and the number of deployed BSs, subject to the constraints on both sensing and communication performance across the corridor. To solve the formulated large-scale nonconvex integer programming problem, we develop a hierarchical coarse-to-fine grid decomposition algorithm. Simulation results demonstrate the benefit of the proposed joint design in reducing the overall deployment cost while ensuring the coverage of the low-altitude ISAC networks.

Integrated sensing and communication is widely acknowledged as a foundational technology for next-generation mobile networks. Compared with monostatic sensing, multi-access point (AP) collaborative sensing endows mobile networks with broader, more accurate, and resilient sensing capabilities, which are critical for diverse location-based sectors. This paper focuses on collaborative sensing in multi-AP networks and proposes a Bayesian probability fusion framework for target parameter estimation using orthogonal frequency-division multiplexing waveform. The framework models multi-AP received signals as probability distributions to capture stochastic observations from channel noise and scattering coefficients. Prior information is then incorporated into the joint probability density function to cast the problem as a constrained maximum a posteriori estimation. To address the high-dimensional optimization, we develop a prior-constrained gradient ascent (PCGA) algorithm that decouples correlated parameters and performs efficient gradient updates guided by the target prior. Theoretical analysis covers optimal fusion weights for global signal-to-noise ratio maximization, PCGA convergence, and the Cramer-Rao lower bound of the estimator, with insights applicable to broader fusion schemes. Extensive numerical simulations and real-world experiments with commercial devices show the framework reduces transmission overhead by 90% versus signal fusion and lowers estimation error by 41% relative to parameter fusion. Notably, field tests achieve submeter accuracy with 50% probability in typical coverage of mmWave APs. These improvements highlight a favorable balance between communication efficiency and estimation accuracy for practical multi-AP sensing deployment. The dataset is released for research purposes and is publicly available at: http://pmldatanet.com.cn/dataapp/multimodal

As is well known, classical PID control is ubiquitous in industrial processes, yet a rigorous and explicit design theory for nonlinear uncertain MIMO second-order systems remains underdeveloped. In this paper we consider a class of such systems with both uncertain dynamics and an unknown but strictly positive input gain, where the nonlinear uncertainty is characterized by bounds on the Jacobian with respect to the state variables. We explicitly construct a three-dimensional region for the PID gains that is sufficient to guarantee global stability and asymptotic tracking of constant references for all nonlinearities satisfying these Jacobian bounds. We then derive a corresponding necessary region, thereby revealing the inherent conservatism required to cope with worst-case uncertainties. Moreover, under additional structural assumptions on the nonlinearities, these sufficient and necessary regions coincide, yielding a precise necessary-and-sufficient characterization of all globally stabilizing PID gains. All these regions are given in closed form and depend only on the prescribed Jacobian bounds and the known lower bound of the input gain, in contrast to many qualitative tuning methods in the literature.

The electrification of off-road heavy equipment presents operational challenges for agencies serving remote sites with limited fixed charging infrastructure. Existing mobile fast charging vehicle (MFCV) planning approaches typically treat fleet design and routing as separate problems, fixing vehicle characteristics before dispatch. This paper formulates a fleet size and mix capacitated vehicle routing problem with time windows (FSMCVRPTW) for MFCV deployment, jointly optimizing fleet composition, charger specifications, routing, and scheduling within a unified mixed-integer linear program. The model incorporates heterogeneous MFCV types with varying power ratings, battery capacities, fuel range, and cost structures, minimizing total daily cost from labor, fuel, amortized capital expenditure, and energy purchase under temporal service windows, resource budgets, and energy-delivery constraints. The formulation is implemented in Python/Gurobi and applied to two case studies using California Department of Transportation wheel-loader data in Los Angeles (dense urban) and Truckee (sparse mountainous). Results show that simultaneous optimization yields compact, well-utilized fleets that meet all service windows while revealing strong sensitivity of unit cost to demand density and geography. The proposed FSMCVRPTW framework provides a generalizable decision-support methodology that co-designs fleet size, charger power, routing, and service schedules in a single optimization layer for context-aware, cost-efficient mobile fast charging.