Papers

The electric power supply for AI data centers is now the most significant bottleneck in the race toward Artificial General Intelligence, surpassing even the constraint of AI accelerator availability. To our knowledge, this paper is the first to describe the end-to-end power management process for a hyper-scale AI datacenter; from early power planning to accommodate next-generation accelerators 6–12 months before their general availability, to tuning power settings after large scale deployment, and finally to dynamic, runtime power management for evolving workloads. We present detailed power measurements for a 150 MW datacenter hosting a cluster of 83K GB200 GPUs. We share insights from building this state-of-the-art AI cluster. We hope this work encourages practitioners across the industry to share their own experiences as well.

Urban heat islands (UHI) are formed due to complex interactions between various factors. UHI, its contributing factors, and their interaction vary over time and location. Accordingly, understanding the causal relation between UHI and its contributing factors is essential to minimizing its adverse effects on the environment and human health. Here, we proposed a statistical method based on Hotelling's T-square test to analyze this association. The proposed test estimates the UHI trends across different urban districts and compares the UHI contributing factors between the districts with increasing and non-increasing UHI trends. This comparison, if significantly different, can be interpreted as evidence of a causal association between the factor and UHI. This research used the proposed test to analyze the UHI and its contributing factors across 22 municipal districts of Tehran between 2003 and 2021. We examined the time series of weather conditions (measured by precipitation, NDSI, and NDWI), vegetation cover (measured by NDVI and EVI), and urban density (measured by NDBI) as factors contributing to the UHI, which was measured through nighttime LST. The results showed that all districts in Tehran exhibited stable or increasing trends in LST, leading to UHI effects. The proposed test indicated that the temporal changes in NDWI and NDBI did not have a causal relationship with UHIs. Meanwhile, variations in other factors were identified as contributing to the intensification of UHIs.

Data streams in real-world industrial scenarios often contain transitional operating conditions that are uncovered during offline training, leading to significant distribution shifts. To bridge the gap between static offline models and dynamic online data, a novel asymmetric adaptation-based fault diagnosis method is proposed in this paper. Specifically, in the offline stage, we employ domain generalization techniques to extract domain-invariant features from multiple stable conditions and construct robust normalized fault prototypes as reference anchors. Subsequently, during online inference, we design an online test-time adaptation method based on a periodic prototype re-projection mechanism to dynamically update prototype positions. Furthermore, we utilize the geometric distribution derived from anchors to guide the updates of classifiers and adopt an asymmetric learning rate strategy for the feature extractor and classifier. The proposed approach ensures rapid adaptation to new transitional conditions while preserving the discriminative power inherited from the offline domain generalization initialization. Experimental results demonstrate that this mechanism effectively leverages offline generalized knowledge to guide online inference, significantly improving robustness in non-stationary environments.

This paper investigates satellite navigation and communication systems in both low-Earth-orbit (LEO) and medium-Earth-orbit (MEO) satellites, which systematically outlines the fundamental principles of satellite navigation systems (SNS), satellite communication systems (SCS), and integrated navigation and communication (INAC) systems. By exploring the enhanced capabilities of satellite systems, the article emphasizes how INAC systems improve overall functionality by enabling efficient signal multiplexing and multiple access, positioning multi-functional satellites as promising alternatives to traditional architectures. Moreover, it introduces emerging frontiers for LEO-based SNS and MEO-based SCS through the integration of advanced sixth-generation (6G) wireless technologies, which cannot be realized through mere extensions of existing communication or navigation techniques. Motivated by these insights, the article further discusses various conceptual transitions required to unlock the full potential of INAC systems, with particular focus on channel capacity, positioning accuracy, and artificial intelligence-enabled waveform design.

In this paper, we consider a mixed ensemble containing a mixture of cesium-type and hydrogen maser-type atomic clocks. For the mixed ensemble, the conventional Kalman filtering algorithm has certain limitations due to divergence of the error covariance matrix. To overcome these limitations, we obtain a Kalman filtering algorithm based on observable canonical decomposition that does not have any diverging terms. We use the estimates from the transformed Kalman filter to propose a time scale generation algorithm called explicit ensemble mean synchronization algorithm for the mixed ensemble. In this algorithm, we synchronize the time deviation of each clock from the ideal clock behavior to the unobservable ensemble mean of the phases where the weighting can be decided by the user. By regulating the free-running dynamics associated with the unobservable state, through choosing an appropriate weight vector, the frequency stability of the generated time scale or the synchronized time shared by the clocks is optimized over shorter (resp. longer) intervals, as measured by Hadamard variance. An illustrative example is given to demonstrate the efficiency of our algorithm.

We present a finite-time framework for identifying stable and unstable linear time-invariant (LTI) systems from a single closed-loop input-output trajectory. The method does not require knowledge of the stabilizing controller, an intermediate observer, or prior separation of the plant into stable and unstable components. The approach uses a non-causal finite impulse response (FIR) model obtained from a Laurent expansion of the transfer function. In this representation, stable dynamics are captured by causal Markov parameters, while unstable dynamics are captured by non-causal coefficients associated with reverse-time stable evolution. This avoids the growth of causal unstable Markov parameters. A key advantage is that the coefficients multiplying both the input and the process noise remain controlled by stable and reverse-time stable decay rates, rather than by growing forward-time unstable dynamics. To handle closed-loop data, we use the injected excitation as an instrumental variable, which removes the bias caused by correlation between the feedback input and the process noise. Under explicit instrument-strength and closed-loop concentration conditions, we derive a non-asymptotic error bound for the estimated Laurent/FIR Markov parameters with the usual $\mathcal{O}(N^{-1/2})$ statistical rate, up to logarithmic factors and truncation terms. The bound captures the effects of process noise, measurement noise, FIR horizons, closed-loop state moments, and controller-dependent instrument conditioning. Numerical experiments support the finite-time analysis by showing the predicted Markov-parameter convergence rate and illustrating how controller-dependent instrument conditioning affects the sample complexity of closed-loop identification.

With the rapid proliferation of wireless and Internet of Things (IoT) devices, ensuring secure and reliable device identification has become a significant challenge. Traditional security techniques, such as IP or MAC address-based authentication, are susceptible to spoofing, whereas Radio Frequency Fingerprint Identification (RFFI) offers a more secure alternative by exploiting the unique hardware imperfections in devices' RF signals. In this paper, we propose a novel deep learning-based framework for RFFI that enhances both accuracy and reliability in challenging RF environments. The core of our approach is the Signal Inception Transformer (SinFormer), which leverages a specialized multi-scale self-attention mechanism to effectively capture both large-scale and fine-grained fingerprints in signals, significantly improving identification accuracy. To further enhance robustness and reliability, we introduce a two-stage training strategy that enables the model to learn general signal features and maintain performance under adverse conditions, such as low Signal-to-Noise Ratio (SNR) or channel variations. The effectiveness of the proposed method is validated using a real-world dataset. Experimental results show that the SinFormer framework consistently outperforms existing methods in accuracy and robustness across diverse and challenging scenarios.

Digital audio broadcasting plus (DAB+) is an attractive illuminator for passive radar because it provides persistent, high-power, and geographically widespread very high frequency (VHF) orthogonal frequency-division multiplexing (OFDM) signals. A channel state information (CSI) sensing approach can convert a single received DAB+ stream into a CSI sequence for radar sensing, avoiding the need for a separately received reference signal in conventional passive radars. However, CSI estimation in DAB+ is challenging due to the differentially encoded communication symbols across time. A wrong symbol transition estimation leads to a persistent multiplicative error in the sequential CSI sequence within a DAB+ frame. This paper formulates single-stream DAB+ passive radar as a posterior-probability-aware differential CSI tracking problem. The proposed method uses the previously tracked CSI as a channel prior, performs prediction-aided maximum a posteriori detection of current symbol, converts posterior transition reliability into observation uncertainty, and applies linear minimum mean squared error fusion to obtain a stable tracking CSI. A reliability-informed CSI fusion strategy is also introduced to preserve weak target information. Theoretical analysis is provided, showing guaranteed performance again in symbol and CSI estimation. Simulation results show that the proposed method can reduce CSI estimation error by over 15~dB compared with prior art. It also improves median target-to-background ratio by more than 11~dB in random fading scenes. Experiments in Sydney, Australia demonstrate improved range-Doppler maps for commercial aircraft sensing.

The global push for electric vehicles (EVs) has sharply increased demand for critical minerals such as cobalt and lithium, creating a tension between rapid industrial growth and long-term sustainability. Extraction is concentrated in a few regions – notably the Democratic Republic of Congo (DRC), Chile, and Argentina – where it has produced serious socio-environmental harms, including ecosystem degradation, labour exploitation, and the displacement of Indigenous communities. In the DRC, cobalt mining is frequently linked to child labour and hazardous working conditions; in Chile, lithium extraction intensifies water scarcity and threatens local agriculture and biodiversity. Policy instruments such as the U.S. Inflation Reduction Act (IRA) seek to promote ethical sourcing, but an extraction-driven model continues to deepen global inequalities. This chapter examines the contested temporalities of the transition, in which the short-term economic incentives of extraction conflict with longer-term environmental and social goals. It argues for a place-based framework built on community-centred governance, sustainable mining practices, and circular-economy strategies, including recycling and material substitution, to align resource security with equity and ensure that the shift to EVs does not reproduce the injustices it aims to address.

Low-density EEG is more suitable for wearable and IoT-based brain sensing, but sparse electrode sampling often lacks sufficient spatial information to characterize cross-regional neural activity. EEG spatial super-resolution aims to recover dense-channel EEG from sparse recordings, yet remains challenging because channel missingness typically occurs at the whole-channel level, spatiotemporal dependencies over the full electrode layout are often underexplored, and the mapping from sparse to dense signals is inherently ambiguous. To address these issues, we propose TGSD, a topology-guided state-space diffusion framework for EEG spatial super-resolution. TGSD first employs a Hierarchical Spatial Prior Encoder to learn topology-aware priors over the complete electrode layout by integrating local geometric relationships with region-level contextual information. Based on these priors and sparse observations, a Conditional State-Space Diffusion Reconstructor progressively generates missing-channel signals through reverse diffusion, while alternating temporal and channel-wise state-space modeling captures long-range temporal dynamics and inter-channel dependencies in a unified framework. Experiments on the SEED and PhysioNet MM/I datasets show that TGSD consistently outperforms representative baselines under different super-resolution factors in both reconstruction fidelity and downstream classification performance. These results demonstrate the effectiveness of combining topology-aware spatial priors with conditional diffusion for enhancing practical low-density EEG sensing in wearable and IoT scenarios. The official implementation code is available at https://github.com/jtggz/TGSD.

This paper presents a proof-of-concept system for localising ground-based WiFi access points, acting as IEEE~802.11mc Fine Time Measurement (FTM) responders, from an uncrewed aerial vehicle using FTM ranging and Global Navigation Satellite System (GNSS)-referenced moving-baseline multilateration. Each associated GNSS-referenced FTM-initiator pose supplies a known reference point, turning the flight trajectory into a temporal multilateration problem. The real-time smartphone pipeline performs GNSS–ranging time association, robust outlier gating, a two-stage Gauss-Newton bootstrap, and sequential Bayesian filtering with bias tracking. Six measurement-noise configurations, including empirical and adaptive models, are evaluated on field data collected in unstructured, mountainous terrain. For a line-of-sight access point with \num{455} ranging measurements, the online Android pipeline achieves a final horizontal error of \SI{4.4}{\metre}, while offline replay of the same flight yields a time-weighted mean horizontal error of \SI{4.7}{\metre} and a best-case final horizontal error of \SI{1.1}{\metre} under the best noise model after a close flyby. For non-line-of-sight targets, the real-time pipeline does not converge because of limited measurement availability, weak geometry, and signal attenuation, although an offline robust least-squares solver recovers a coarse solution for the vegetation-only case. The system is intended as a building block for Networked One Search Agent architectures, and preliminary middleware tests demonstrate software-level interoperability, while quantitative multi-agent accuracy is left for future work.

Low-power wireless-capable systems-on-chips (SoCs) are critical for researching many of our current environmental issues. The scale at which these devices are needed for many applications necessitates innovation in their design to reduce the various capital and labor costs involved with operating an extensive sensor network. This can be difficult for devices with novel wireless architectures, as many emerging architectures lack commercially available development platforms. This makes pre-silicon validation challenging, and the impact of a failed tapeout is unacceptable when the cost is of primary concern for these devices. In this work, we propose a digital twin ecosystem for Bluetooth Low-Energy (BLE) with physical-layer (PHY) control intended for novel device development and demonstrated through use with crystal-free single-chip sensor motes. We present this system operating with multiple RF front ends and digital baseband implementations, including a commercially available Software Defined Radio (SDR) with synthesized RTL and embedded firmware, along with an existing crystal-free SoC front end and FPGA digital baseband. These configurations are shown to be capable of communicating sensor data with commercially available BLE devices and achieving receiver sensitivities up to -82 dBm, exceeding the minimum BLE specification. This approach is extendable to other hardware and communication protocols and promises to enable inexpensive, reusable validation and verification tools for novel wireless devices.

Many chemical engineering systems are governed by mechanisms that switch across operating regimes, making the data-driven discovery of regime-dependent governing equations essential for predictive modeling, optimization, and control. We propose symbolic decision trees for the data-driven discovery of regime-dependent governing equations. The method simultaneously learns interpretable splitting conditions to partition the input domain and local governing equations that describe each regime. To improve tractability, both the splitting conditions and governing equations are parametrized using basis functions, resulting in a mixed-integer optimization learning problem. We use the proposed approach to learn hybrid dynamical models and a constitutive equation for the zero-shear viscosity of polymer melts. Symbolic decision trees identify physically interpretable regimes and local governing equations while improving predictive accuracy relative to approaches that learn a single global model or use existing decision tree models. This framework provides an interpretable and generalizable route for discovering regime-dependent models in chemical engineering systems.

A critical challenge facing clinicians managing chronic disease interventions is sustaining long-run patient health given limited information and resources. Digital therapeutics (DTs) provide a cost-effective way to manage interventions at scale through repeated interactions (e.g. daily treatment recommendations), but patient success is highly dependent on their adherence. Behavioral psychology suggests that both treatment recommendations and past adherence affect future adherence, yet existing decision support frameworks for DTs model only recommendation effects or treat adherence as exogenous context, leaving a key gap in model and algorithm development. To address this gap, we present a DT decision support framework that captures both recommendation and adherence effects, allowing clinicians to better plan treatment recommendations. We model a patient's time-varying capacity for engagement with treatment using a linear dynamical system (LDS) that captures both recommendation and adherence effects, endogenously connected to adherence behavior with a logit link. We establish finite-time identification guarantees for this model, extending LDS results to our setting. Next, we propose an optimism-based algorithm, UCB-BOLD, for online treatment selection and prove that it achieves sublinear regret. We evaluate UCB-BOLD against benchmarks via ablation studies on a synthetic patient cohort generated using micro-randomized trial data. DT decision support tools can include dynamical models to enable decision makers to efficiently use the data in DT settings to improve patient health through effective resource allocation. While myopic or heuristic approaches suffice for some patient types, the benefits of explicitly planning around recommendation and adherence effects are significant for others; UCB-BOLD achieves 2-3x lower conditional value-at-risk regret than the next-best benchmark.

This article introduces the "chart & chirp" method of one-shot, in situ VCO tuning curve estimation, learning, and predistortion, which provides an alternative to prevailing LMS-based background calibration loops. The proposed approach utilizes a cycle-counting FDC to estimate the VCO tuning curve and a linearly interpolating QDAC to actuate the predistortion. An LLSE-based learning process generates predistorted DAC control codes for any physically achievable chirp parameters. In simulation, 50 and 51 kHz of RMS FM error is achieved for 5 and 20 us chirps, respectively. The locations of nonlinearity-induced spurs in the IF spectrum are predicted via analysis of the DFT of FM error in generated chirps, with higher frequency DAC updates weakening these spurs and moving them out of the IF band of interest. Simulation in a coherent, monostatic radar model reveals that deterministic phase error from the chirp nonlinearity arises at a level above -70 dBc/Hz at 1 MHz offset, dominating random phase noise and setting the IF SNDR for R < 23 m and R = 92 m around 40 dB and 24 dB, respectively.

This study explores the use of Visible Light Communication (VLC) in Collective Perception (CP), a technology that enables vehicles and infrastructure to share sensor information to help reduce traffic accidents. Recent advances in Vehicle-to-Everything (V2X) communication have spurred growing research interest in CP. However, in regions such as the United States and Japan, only 30 MHz of radio spectrum is allocated for V2X, which is insufficient to effectively support CP. In this paper, we propose integrating VLC into V2X systems to enhance CP, complementing the existing 5.9 GHz band for V2X communications. VLC can coexist with wireless systems that use radio waves, providing an additional optical channel for data exchange. To the best of the authors' knowledge, this is the first study to investigate VLC for CP. We evaluate the feasibility of VLC-based CP through three experiments. First, we measured the application-level delay of a VLC-based CP system in a stationary indoor environment. Next, we evaluated its communication range in a stationary outdoor setting. Finally, to assess robustness under realistic conditions, we conducted driving experiments at vehicle speeds up to 90 km/h. The results demonstrate that VLC-based CP is feasible and could serve as a promising solution to spectrum scarcity in the 5.9 GHz band for future V2X communications.

Preserving stability is a central problem in data-driven model order reduction of dynamical systems. For linear systems whose dynamics depend on geometric or physical parameters, multivariate rational approximation algorithms such as the Parameterized Sanathanan-Koerner iteration and the pAAA algorithm construct parameterized reduced models from sampled transfer function data. In this setting, stability must be enforced robustly across the parameter domain. This paper introduces a necessary and sufficient criterion for characterizing the stability of parameterized models. Within a unified framework, the results apply to models with general rational as well as polynomial dependence on the parameters. Building on this criterion, we develop and demonstrate a rational approximation algorithm that includes robust stability constraints through convex optimization. Relative to the state of the art, the approach enforces stability without conservatism while allowing increased flexibility in the choice of model structure.

Frequency-modulated continuous-wave radar sensing often relies on labeled measurements that are costly, restricted, or difficult to collect at scale. This work evaluates physics-informed digital twins as controlled testbeds for early-stage quantum-classical radar learning. Two synthetic radar benchmarks are considered: unmanned aerial vehicle classification from range-Doppler maps and human fall detection from Doppler-time spectrograms. For both tasks, inputs are standardized, reduced using principal component analysis, and classified using either a radial basis function support vector classifier or a quantum support vector classifier. All quantum-kernel results are obtained using noiseless classical simulation; no quantum hardware is used, and no quantum-advantage claim is made. Across five random seeds, the quantum support vector classifier improves the UAV benchmark from four principal components onward, reaching an accuracy of 0.941 +/- 0.012 at eight components, compared with 0.880 +/- 0.029 for the classical baseline. On the fall-detection benchmark, both classifiers perform similarly, with a small quantum-kernel improvement at higher feature dimensions. A Gaussian-noise robustness study shows limited performance degradation across the tested noise levels, while preserving the UAV quantum-kernel gain. These results support digital twins as useful, controlled environments for radar-QML benchmarking prior to measured-data validation and hardware execution.

Parkinson's disease (PD) is a highly heterogeneous disease, including which motor symptoms are dominating. Imaging biomarkers that support subtype stratification could also improve biological understanding and study design, and enable personalized treatment strategies. This study evaluates whether deep-learning based automatic brain segmentation, in addition to quantitative maps from 7 Tesla MRI, can highlight differences between Healthy Controls (HC), Postural Instability and Gait Difficulty (PIGD) and Tremor Dominant (TD), and subsequently be used for objective PD stratification. The performance of machine learning classifiers may be improved with feature selection. 21 HC, and 24 people with PD (PwP) were included. The U-Net training was assessed with DSC. Two classification approaches using 5-fold cross-validation were defined across three tasks: (1) HC vs PwP; (2) PIGD vs TD; (3) multiclass, HC vs PIGD vs TD. Approach A used all extracted features. Approach B found the optimal subset of features for the classification tasks. The U-Net achieved mean DSC of 0.86 for all ROIs during training. Approach A: Task 1 best accuracy of 0.69 and best AUC of 0.73. Task 2 accuracy 0.69, AUC 0.90. Task 3 accuracy 0.62, AUC 0.66. Approach B: Task 1 accuracy of 0.82 and AUC of 0.93. Task 2 accuracy 1.00, AUC 1.00. Task 3 accuracy 0.73, AUC 0.91. DL-based segmentation combined with qMRI feature selection improved classification relative to using all features, supporting the potential of interpretable, low-dimensional imaging signatures for PD diagnosis support and phenotype stratification. Larger, multi-site studies are warranted to assess generalizability and stability.

This paper provides a comparative study of modern uncertainty quantification (UQ) methods. To greatly enhance real-time performance, both differential algebra (DA) and a directional differential algebra (DDA) approach are employed. This can enable fast UQ in the case of non-Gaussian statistics. Higher-order moments, namely skew and kurtosis, can be computed quickly by several means. This motivates their implementation in an analytic approximation of the confidence bounds for the so-called "banana-shaped" non-Gaussian distributions encountered often in nonlinear astrodynamics problems. This method improves greatly on a linear covariance approach, with only 5x its runtime in numerical tests, even before DA methods are employed. Test problems in this work include a restricted three-body cislunar example and an Earth-return aerocapture example.

This paper addresses the problem of reaching consensus under input saturation and intermittent communication, which can hinder the convergence of the system. We propose a method that translates the consensus into an equivalent stability problem. Then, we compute bounded sets that enclose the initial conditions and the evolution of trajectories leading to local input-to-state stability for systems interconnected over directed intermittent topologies. Our contributions include sufficient conditions for stability and stabilization of multi-agent systems under intermittent interactions and saturating inputs, with the ability to evaluate disturbance tolerance and rejection based on the regions that enclose the system's trajectories. We define disturbance rejection in terms of the $\mathscr{L}_2$ gain, and formulate stability and controller design conditions as convex optimization problems. Our method enable the maximization of regions that ensure local input-to-state stability, we provide numerical examples highlighting the trade-offs between mean frequency of intermittent interactions, disturbance energy, and convergence region size.

Machine learning deployments in real-world wireless communication tasks face significant generalization challenges due to location and environment-specific signal structure, high diversity in data across different deployments, and limited availability of real-world data. Current approaches for assessing data similarity between training and inference (deployment) distributions, as well as evaluating model transferability, suffer from high computational costs and inconsistent performance, leaving critical model deployment and model life cycle management decisions without a principled foundation. To address this, we introduce a dataset similarity framework built upon the feature space of a pretrained wireless foundation model. Our method, LWM-CDE (Contrastive learning of Dataset Embedding), fine-tunes the dataset embeddings of the foundation model using a combination of contrastive and geometry-shaping losses, creating a structured manifold where distance reliably indicates transferability. Extensive experiments on wireless benchmarks show that LWM-CDE achieves stronger correlation with empirical transfer performance than existing metrics while being more computationally efficient. The learned representation space supports more effective and data-efficient decision-making for tasks like source dataset selection, label-aware augmentation, and budgeted pretraining, demonstrating its broader utility across different wireless communication applications.

In industrial scenarios involving multi-agent collective decision-making, centralized decision-making may not be admissible due to restrictive access to individual local information, while the conflicts between participants' self-interest and global performance may also impede collaborative distributed decision-making. This paper proposes a mechanism-driven distributed decision-making method, wherein incentives are employed and designed to motivate participants to collaborate in a distributed fashion even though each participant's decision is driven primarily by self-interest. Focusing on optimization problems with coupled objective functions and coupled constraints, we design a distributed optimization algorithm tailored for this class of problems and provide guarantees for its convergence. Furthermore, we design two incentive mechanisms, the shadow pricing mechanism and the Vickrey-Clarke-Groves mechanism, and demonstrate that participants are willing to engage in distributed collaboration under these mechanisms. The mechanism drives the execution of the distributed algorithm, and the optimal result of distributed computation guides the determination of incentives in the mechanism, both of which are interrelated to form a closed loop. Finally, numerical experiments illustrate the effectiveness of the proposed algorithm and mechanisms.

In this technical report, we describe our submission for the WildSpoof Challenge TTS Track: Text-to-Speech with In-the-Wild Data. We introduce F5-TTS-DPS, a model built upon the F5-TTS architecture. Our approach integrates Exponential Moving Average (EMA) into supervised fine-tuning to stabilize training and improve generalization. To enhance synthesis fidelity, we leverage large language models (LLMs) and large audio language models (LALMs) for dual-scoring prompt selection, filtering reference audio and text prompts to ensure quality while addressing alignment issues in noisy datasets. Experimental evaluation demonstrates that F5-TTS-DPS achieves strong performance with UTMOS of 3.20 and speaker similarity of 0.51 on the development set. More importantly, our model achieves the best a-DCF scores of 0.1582, 0.5233, and 0.2562 across three advanced SASV systems among all submissions, indicating our synthesized speech is the most difficult to detect and exhibits the highest degree of naturalness and authenticity. Combined with competitive WER performance, these results validate the effectiveness of our approach in generating natural-sounding speech with strong spoofing capabilities.

A surrogate-based synthesis framework for antenna arrays is presented that incorporates mutual coupling while keeping optimization computationally efficient. The method combines a common characteristic-mode basis, a global modal coupling model, and element-wise generalized scattering matrices (GSMs). Array design variables are formulated and optimized on physically meaningful manifolds, in particular the manifold of unitary symmetric matrices for reciprocal and lossless element GSMs. A staged penalty strategy is used to progressively enforce sidelobe and cross-polarization constraints during multi-beam optimization. The framework is demonstrated for an 8x8 left-handed circularly polarized patch phased array with scan behavior in one principal plane. Different degree-of-freedom assignment strategies are compared, showing that constrained non-identical element classes can satisfy stringent pattern requirements where equal-element designs fail. For the demonstrated case, the optimization converges within seconds on a single CPU core, and full-wave verification of the realized arrays confirms the predicted trends, with good agreement for the SLL and useful accuracy for the XPR. The results indicate that the proposed formulation is a practical and scalable route for coupling-aware array synthesis and realization.

3rd Generation Partnership Project (3GPP) Technical Report (TR) 38.901 channel models (Releases 15-19) are widely used for physical-layer design and system-level evaluation in dense urban outdoor-to-indoor (O2I) and indoor environments. These models capture ensemble-averaged channel statistics but do not account for site-specific geometry. In this paper, we compare Power Delay Profiles (PDPs) derived from a deterministic ray-tracing model (Remcom Wireless InSite software) with those from the 3GPP TR 38.901 Tapped Delay Line (TDL) channel models. This comparative analysis is performed using a dense urban O2I scenario and a representative single-story indoor layout modeled in Washington, D.C., under matched link-distance and Non-Line-of-Sight (NLOS) conditions. All Wireless InSite PDPs are power-normalized to enable comparison of relative multipath delay structure. We evaluate root-mean-square (RMS) delay spread, mean excess delay, effective maximum delay, and Kullback-Leibler (KL) distribution divergence. Results indicate that 3GPP TDL models generally exhibit longer delay spreads and often fail to capture deterministic, site-specific features such as late-arriving energy and irregular spikes. While TDL models can approximate primary channel features in some cases, their reliance on ensemble-averaged statistics rather than geometry limits their representation of fine multipath structures. We conclude that while 3GPP TDL models are suitable for large-scale system evaluation, deterministic or hybrid approaches are more appropriate for site-specific physical-layer design.

Modern optimal control theory involves adjoining the already known equations of motion of a dynamic system to the objective function using dynamic costates; this is done in order to constrain the optimal control solutions to satisfy the equations of motion. The use of costates increases the number of variables and hence increases the complexity of the problem. On the other hand, variational methods of analytical mechanics finds the equations of motion by minimizing an action functional of the dynamic system, realizing control forces as external input to the system. In this paper a new disruptive approach for computing the optimal control is presented. This approach adopts the variational methods of analytical mechanics to derive equations for the control, in addition to the equations of motion. This is achieved by recognizing the control actuator as part of the dynamic system. In addition to the kinetic energy and potential energy, the action functional in this new approach includes additional energy terms that represent the control energy of the system. Two different methods are presented to write the modified action functional. The proposed approach is a significant departure from the modern optimal control theory, and it eliminates the need for costates when solving for the control. In this paper, a case study is presented to demonstrate the new approach.

This paper presents a system design methodology for a large-scale static wireless mesh network for 155 commercial off-the-shelf (COTS) radio nodes at fixed infrastructure sites in a challenging island environment. The architecture consists of approximately ten 15-node clusters, each with designated primary and secondary gateway nodes to support inter-cluster communication. A structured, multi-stage planning methodology was developed to guide network design. Site-specific radio frequency (RF) path loss predictions were generated using Remcom's Wireless InSite ray-tracing platform, incorporating terrain, buildings, and dense foliage effects. To optimize connectivity under physical-layer and operational constraints, spectral embedding combined with balanced k-means clustering was applied to partition the nodes into clusters of comparable size. A link budget analysis determined the maximum tolerable path loss under waveform and hardware constraints, defining the connectivity threshold used in the clustering framework. This work integrates deterministic RF propagation modeling with constrained clustering optimization to provide a scalable framework for planning static wireless mesh networks in complex geographic environments. Node mobility and higher-layer networking protocols were outside the scope of this study.

The Open Radio Access Network (O-RAN) architecture allows AI to be embedded directly into the RAN through modular xApps and rApps, yet creating these applications collecting data, training models, writing code, and deploying them safely remains slow and largely manual. Large Language Models (LLMs) offer strong reasoning and code-generation capabilities but are unsuited for the fast, deterministic inference required in real-time RAN control. We present a proof-of-concept Dual-Brain architecture that combines both strengths: an LLM-based orchestrator translates operator intents into data-collection policies and deployment code, while an automated ML engine, NeuralSmith, trains lightweight classifiers on demand via an API. We describe the architecture and provisioning workflow, share practical insights from a containerized O-RAN 5G~SA testbed, and discuss open research directions.

Modern audio-visual media rely on compact representations for efficient storage and transmission, whereas realistic digital touch still depends on high-resolution tactile recordings. Existing approaches for representing tactile signals constrain manipulation and limit the generation of new content. Here, we introduce two compact representations, spectral beta and spectral slope, that capture the temporal spectral structure of finger-surface friction signals while preserving perceptually relevant information. Spectral beta models spectral skewness using a two-parameter beta distribution, whereas spectral slope approximates the spectrum with an asymmetric bandpass filter defined by low- and high-pass orders. We evaluated these representations in a perceptual study with 14 participants using five virtual textures rendered on a friction-modulation display and compared them with physical textures and high-fidelity reproductions of recorded signals. Spectral beta achieved perceptual similarity ratings comparable to those of the original high-fidelity reproductions. Regression analysis further showed that matching spectral energy across nine critical frequency bands was the strongest predictor of perceived realism. Together, these findings suggest that tactile texture perception depends primarily on fundamental temporal spectral patterns and that modeling these patterns is sufficient for perceptually realistic rendering. These results establish an efficient and scalable framework for haptic compression, communication, and synthetic texture generation.