Papers

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 (Ventio), 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 Ventio. 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.

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.

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 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: https://github.com/syu-kylin/HRVConformer.

We evaluate the M2M4 and EVM methods for real-time SNR estimation in FSO communication systems subject to deep fading. Using an experimental setup with controlled deep fading, we show that the M2M4 estimator reliably tracks the SNR profile, making it suitable for triggering transceiver adaptation.

This paper presents a robust tracking controller for tracking curvature-constrained paths by vehicles/robots with uncertain Dubins dynamics. Although Dubins paths have been widely used in vehicular and robotic applications, robust and convergent tracking under model uncertainties remains understudied. To address this, we propose path tracking controllers based on sliding mode control, formulated in the transverse coordinate frame, which guarantee invariance and convergence of both lateral and heading errors to zero in the presence of bounded disturbances. Simulation results show that the proposed method reliably tracks paths despite disturbances and significantly outperforms existing methods based on sliding mode controllers.

Electromagnetic inverse scattering imaging is a crucial computational technique in microwave sensing systems. However, quantitative reconstruction of high-contrast targets remains a formidable challenge due to severe multiple scattering effects. To overcome this bottleneck, this communication proposes an alternating optimization framework based on a cross-correlated implicit neural representation (Alt-CC-INR) as a preliminary study towards the robust solution of high-contrast targets reconstruction. Unlike conventional purely neural architectures, this framework introduces a heterogeneous optimization paradigm. Specifically, the method employs an analytical Polak-Ribière conjugate gradient (PR-CG) algorithm, driven by a cross-correlated loss and accelerated by 2D-FFT, to optimally resolve the dynamic contrast sources. Subsequently, a stochastic deep learning optimizer updates the continuous neural representation to infer the static medium profile. Validations on simulated and measured data demonstrate that Alt-CC-INR effectively overcomes local minima when inverting high-contrast objects, providing a highly robust computational electromagnetic solver.

Recent video super-resolution (VSR) approaches use deep neural networks to enhance low-quality input videos and recover visual detail, with diffusion-based methods in particular showing promising results. In this paper, we investigate whether existing video quality models can be used to assess the performance of these diffusion-based VSR methods, by comparing model predictions with results from a subjective test. The study compares six upscaling methods (Lanczos, Rhea, SCST, DOVE, SeedVR2, Starlight Mini) applied to both compressed (AV1 and DCVC-RT) and uncompressed low-resolution videos considering the play-out on a UHD-1/4K screen. A range of full- and no-reference quality models are used to assess their applicability to this new type of quality degradation, focusing on within-sequence performance. The results highlight that CNN-based full-reference models, such as LPIPS, DISTS, and CVQA-FR show significantly higher correlation coefficients than both conventional full- as well as the tested no-reference models. Most overestimate the overly sharp results of SCST, with VMAF mainly failing due to spatial inconsistencies introduced by Starlight Mini. None of the tested video quality models reach sufficient accuracy so as to replace complementary subjective testing. The reference, degraded and upscaled videos, as well as the user ratings and model scores are made available with the paper at https://github.com/Telecommunication-Telemedia-Assessment/AVT-VQDB-UHD-1-VSR as open data.

We describe the winning system for Task 2 of the KSAA-2026 Shared Task on Arabic Speech Dictation with Automatic Diacritization. The task requires producing fully diacritized Arabic text from speech audio and undiacritized transcripts, with only 2,327 training samples available and no external data permitted. Our system fine-tunes CATT-Whisper, a character-level multimodal model combining a pretrained CATT text encoder with a frozen Whisper speech encoder. The key to our approach is training regularization: R-Drop consistency regularization, Optuna-optimized hyperparameters with high weight decay, and Focal Loss. At inference, we average 200 stochastic forward passes across four model checkpoints using Monte Carlo Dropout at the softmax probability level. The system achieves 23.26% WER on the primary leaderboard metric (with case endings, including no-diacritic positions), placing 1st among all participants.

Pathological assessment guides lung cancer diagnosis, treatment selection, and prognostic evaluation, yet current CPath approaches rely on task-specific models for isolated objectives. Although pan-cancer foundation models offer versatility, they lack subspecialty-level depth and have not been evaluated across clinical workflows or prospectively validated in real-world settings. We introduce PulmoFoundation, a multi-center, prospectively validated, randomized controlled trial (RCT)-evaluated foundation model for comprehensive lung pathology assessment across pre-operative, intra-operative, and post-operative care. Built upon Virchow2 via subspecialty-specific pretraining using ~40,000 diagnostic H&E-stained whole-slide images (WSIs), PulmoFoundation was systematically evaluated on ~26,000 WSIs across 32 clinically relevant tasks. In addition to accurately predicting molecular markers and patient survival, our model achieves clinical-grade performance in core diagnostic tasks across biopsy, frozen section, and surgical resection slides. In a registered prospective study of 1,357 patients across 11 diagnostic tasks, our model achieved an average AUC of 92.3%. Using pre-specified triage thresholds, PulmoFoundation could reduce additional second-review burden for 68.8% of biopsies and 83.0% of frozen sections, and defer 44.5% of IHC stain orders, with PPVs of 1.000, 0.991, and 0.966. Beyond prospective validation, we conducted a crossover RCT with eight pathologists, in which AI assistance improved diagnostic accuracy across 5,264 case-reader pairs (91.7% w/ AI vs. 83.2% w/o AI). AI assistance also reduced median diagnostic time by 18.3%, increased diagnostic confidence by 9.0%, and improved inter-rater agreement from moderate (kappa = 0.55) to substantial (kappa = 0.76). Together, these evaluations support PulmoFoundation as a clinically validated decision-support system for lung pathology.

The aim of this Lecture Note is to introduce the Signal Processing (SP) community to a powerful yet still under-utilised tool: the semiparametric statistics. In short, the semiparametric framework allows us to estimate or perform hypothesis testing on a finite-dimensional parameter $\boldsymbol{\theta} \in \Theta \subseteq \mathbb{R}^p$ in the presence of an infinite-dimensional nuisance parameter $g \in \mathcal{S}$ (i.e. a function), such as the density of the noise. Clearly, this framework is general enough to include almost every SP application. Remarkably, as the title suggests drawing on George R. R. Martin's famous book series, the greatest advantage of semiparametric statistics over parametric and non-parametric ones lies in the fact that it is able to reconcile two seemingly dichotomous concepts: statistical efficiency and distributional robustness. To explain exactly what this means, in this Lecture Note we will focus our attention on the famous and fundamental symmetric location problem.

Controlling partial differential equations (PDEs) with learning-based policies remains fundamentally limited by fixed-dimensional representations: policies trained for a specific sensor, actuator, or agent configuration typically fail when the configuration changes. This limitation is particularly severe in multi-agent PDE control, where policies do not scale across population sizes without retraining. We address this challenge by introducing Cardinality Invariant Neural Operator Control (CINOC), reformulating PDE control as an operator learning problem that maps state fields to continuous control functions and trains them end-to-end through differentiable PDE solvers, yielding policies that naturally adapt to varying sensor and actuator configurations. Remarkably, CINOC policies trained on small swarms exhibit cardinality invariance, allowing for zero-shot transfer to significantly larger populations as well as robustness to partial agent failure. This scalability arises from agents sharing a common policy and coordinating through their physical environment, which produces an emergent self-normalization effect. To explain this phenomenon, we provide a theorem grounded in mean-field theory demonstrating that policy gradients computed from finite-agent systems converge to those of a continuous control limit. Empirically, we validate CINOC on tracking, stabilization, and density transport across linear, nonlinear, chaotic, and turbulent PDEs.

This paper proposes a coordinated energy-mobility dispatch framework for grid support service provision in smart cities under time constraints. In particular, a scenario in which a distributed system operator requests a specified amount of energy within a given deadline is considered. A fleet of connected autonomous electric vehicles equipped with virtual battery partitioning is dynamically dispatched toward vehicle-to-grid stations. The routing problem is formulated as a periodically updated resource-constrained shortest path, accounting for time and energy constraints with congestion-dependent travel times derived from a dynamic traffic model. At the vehicle level, a model predictive control strategy regulates speed to satisfy mobility energy requirements while ensuring deadline compliance. The framework is validated through simulations on the urban network of Rapallo (Italy), demonstrating robustness against congestion-induced delays.

We investigate deterministic and nonblocking supervisory control of discrete event systems under cyber-attacks using the ALTER (Attack Language for Transition-basEd Replacement) model. While prior works consider supervisory control that achieves either the large (upper bound) language or small (lower bound) language separately, deterministic supervisory control achieves both large language and small language at the same time to ensure that the language generated by the supervised system is unique and deterministic. We introduce two new concepts of CA-D-controllability and CA-D-observability and prove that they are necessary and sufficient for the existence of a deterministic supervisor. For nonblocking supervisory control, the objective is to ensure that the supervised system can always reach marked states under any attack scenario. We prove that relative closure, CA-D-controllability, and CA-D-observability together are necessary and sufficient for the existence of a nonblocking supervisor. We further develop methods to verify CA-D-controllability and CA-D-observability. We also illustrate our results using a robotic system example.

Quantum wireless sensing using Rydberg atomic receivers enables high-sensitivity signal acquisition direction-of-arrival (DoA) estimation. However, it suffers from a fundamental limitation, where only the magnitude of the received signal is observable. The recently proposed Quantum-MUSIC algorithm addresses this problem by recovering phase information through alternating minimization and subsequently applying the MUSIC algorithm for DoA estimation. However, the existing approach relies on an $\ell_2$-norm phase retrieval step, making it highly sensitive to outlier measurements produced by hardware faults, sensor saturation, or adversarial interference. In this letter, we propose a Robust Quantum-MUSIC (RobQMUSIC) framework that replaces the $\ell_2$-norm with an $\ell_1$-norm formulation. The resulting weighted phase-retrieval problem is solved efficiently via an Iteratively Reweighted Least Squares (IRLS) scheme embedded within the alternating minimization loop, requiring no increase in structural complexity relative to the baseline algorithm. Simulation results demonstrate that RobQMUSIC achieves near-identical DoA estimation accuracy to Quantum-MUSIC under ideal conditions, while maintaining robust performance over a wide range of outlier contamination levels at which Quantum-MUSIC fails entirely.

Ultra-low-bitrate speech coding is pivotal for bandwidth-constrained communication and deep compression, yet maintaining naturalness and speaker identity at such extreme bit budgets remains challenging due to pronounced information loss and quantization instability. To this end, we propose FMelCodec, an ultra-low-bitrate neural speech codec in the mel-spectrogram domain, cast as a three-stage coding-refinement-reconstruction (CRR) framework that can operate at as low as 250 bps. In the CRR framework, the front-end mel-spectrogram coding stage employs a highly aggressive 640x compression/decompression encoder-decoder structure with a single 1024-entry VQ codebook, coupled with an online clustering strategy that reassigns underused codewords to prevent codebook collapse and preserve codebook diversity. The subsequent conditional flow matching (CFM)-based mel-spectrogram refinement stage leverages a lightweight velocity-field estimator and CFM-based solver to refine the codec-degraded mel-spectrogram produced by the preceding decoder, and adopts a self-consistency training scheme that supports fewer iterative inference steps for the purpose of reducing computational overhead. Finally, the vocoding-driven waveform reconstruction stage employs a HiFi-GAN vocoder to faithfully reconstruct waveform from the refined mel-spectrogram. Experiments conducted on two datasets spanning two sampling rates show that, under ultra-low-bitrate constraints of 250 bps for 16 kHz and 750 bps for 48 kHz, both objective and subjective evaluations consistently demonstrate that FMelCodec achieves higher speech reconstruction quality and speaker similarity, while incurring lower computational and model complexity.

To date, most of the research on transport planning has focused on optimizing revenues or utilitarian metrics such as average travel times, which often ends up penalizing the worst-off for the sake of profit or efficiency. At the same time, most of the research in transport justice has focused on assessing injustices, without being able to prescribe operational solutions. This paper contributes to bridging this gap and presents optimization models for justice-informed operational planning of intermodal mobility systems that explicitly account for the budget and safety limitations of users, and for infrastructural capacity constraints. Specifically, we first focus on an intermodal Autonomous Mobility-on-Demand (AMoD) system – where self-driving robotaxis provide on-demand mobility jointly with public transit and active modes – and characterize its operations from a mesoscopic planning perspective via network flow models. Second, we leverage these models to optimize system operations through both utilitarian efficiency and justice-informed objectives. We showcase our framework in a real-world case-study for Manhattan, New York. Our results show that monetary budgets significantly limit the social justice potential of AMoD systems if they are to be deployed as transportation network companies. At the same time, granting free public transit can result in sufficiency levels very close to a completely free intermodal AMoD system, where justice-informed operations can be achieved without compromising standard efficiency metrics, ultimately highlighting the strong potential of social policies.

In the past decade, numerous studies have applied deep neural networks (DNNs) to decode auditory attention (AAD) from Electroencephalogram (EEG) signals via stimulus reconstruction. However, the influence of dataset balance on the decoding performance of stimulus reconstruction-based AAD remains unexplored. In this study, three publicly available EEG-AAD datasets - KUL, DTU, and NJU cEEGrid - are used to construct both balanced and unbalanced experimental conditions. We hypothesize and demonstrate that stimulus reconstruction-based DNN decoders tend to produce overestimated decoding performance on unbalanced datasets. To address this issue, we propose a leave-one-paired-envelope-out (LOPEO) cross-validation protocol. Experimental results confirm that LOPEO effectively prevents inflated decoding accuracy on unbalanced datasets. While balanced datasets are generally preferred in experimental design, LOPEO provides a principled evaluation framework for unbalanced datasets that have already been published, filling an important gap in the field.

Integrated sensing and communications (ISAC) is a key use case for sixth-generation (6G) wireless systems, where parametric channel estimation (PCE) plays a central role in enabling sensing, localization, and channel equalization in high-mobility scenarios. However, PCE is typically more computationally demanding than conventional channel estimation, which motivates the development of lower-complexity solutions. In this letter, we propose a fast PCE algorithm for time-varying and frequency-selective (TVFS) channels based on canonical polyadic (CP) decomposition and tensor processing, combined with ESPRIT-based initialization, component refinement, and exact line-search alternating coordinate descent. Two variants are presented: one for fully digital and another for hybrid receiver architectures. Numerical results show that the proposed method clearly outperforms a related CP-based baseline while achieving estimation performance close to a multiple-start SAGE benchmark at a substantially lower computational cost, with about one order of magnitude shorter execution time.

Motivated by structural biology applications, we study the projected multi-reference alignment (MRA) model, in which an unknown signal is observed through noisy samples, each generated by applying a random cyclic shift followed by a fixed projection. The projection merges reflection-symmetric index pairs, thereby discarding orientation information. The goal is to recover the dihedral orbit of the signal. We prove that in the high-noise regime, the first three moments of the projected observations determine a generic dihedral orbit. The main mechanism is a reduction, at the moment level, from projected MRA to the reflection-invariant phase-coupling structure of dihedral MRA. In Fourier-cosine coordinates adapted to the projection, the first moment determines the mean component, the second moment determines the Fourier magnitudes, and selected third moments yield the cosine phase-coupling relations appearing in the dihedral bispectrum. These relations lead to a constructive recovery scheme from moments up to order three. We complement the population theory with finite-sample experiments comparing expectation–maximization (EM), direct moment optimization, and direct Fourier-cosine moment optimization. The results show that, in the high-noise regime, both EM and direct moment optimization are consistent with the predicted third-moment sample-complexity scaling $n \gtrsim \sigma^6$, where $n$ is the number of observations and $\sigma^2$ is the noise variance.

Deep reinforcement learning (DRL) has long been a promising solution for sequential resource management in wireless networks. However, conventional DRL methods are fundamentally limited by their reliance on unimodal policy distributions, inefficient exploration in high-dimensional action spaces, and poor adaptability to dynamic and heterogeneous environments. Meanwhile, diffusion models (DMs) as one of the most powerful families of generative AI have demonstrted remarkable capabilities in modeling complex, multi-modal data distributions across diverse domains. The integration of DMs and DRL has opened a new and rapidly growing research direction, in which DM-enabled policies substantially enhance decision quality by capturing the complex, discontinuous, and multimodal action structures inherent in wireless resource management. In this paper, we present a comprehensive survey of DM-enabled DRL algorithms and their applications for various issues in wireless networks. Particularly, we first provide the theoretical background of DM and present different DM-enabled DRL algorithms. We then systematically review applications of DM-enabled DRL for across computation offloading in mobile edge computing, UAV-assisted, vehicular, and AIGC-driven systems, as well as wireless resource allocation, physical-layer security, and robotics and UAV planning. We conclude the paper by higlight future research directions.

Directional-statistics-based mask-based blind speech separation (BSS) clusters normalized time-frequency (TF) observations from $M$ microphones on the complex unit sphere, without relying on plane-wave or spherical-wave assumptions. Existing methods use separately defined angular mixture models, which makes the effect of density shape difficult to isolate. This paper proposes a complex spherical Student's $t$ mixture model (cSTMM) that connects the complex angular central Gaussian mixture model (cACGMM), complex Bingham mixture model (cBMM), and complex Watson mixture model (cWMM) through the degrees of freedom $\nu$ and eigenvalue constraints. We derive a latent-scale expectation-maximization (EM) framework with an approximate M-step based on high-concentration approximation (HCA). On noise-free LibriSpeech mixtures reverberated using measured room impulse responses (RIRs), the development-selected value $\nu^\ast=1$ outperformed the cACGMM-equivalent choice $\nu=M$ in all 18 test conditions, yielding a mean signal-to-distortion ratio improvement (SDRi) gain of $0.25\,\mathrm{dB}$. The model reduces to the cACGMM at $\nu=M$ and approaches the cBMM/cWMM in the large-$\nu$ limits.

Most neural vocoders are limited to one type: either GAN or diffusion-based. While state-of-the-art models like Vocos and WaveNeXt use powerful ConvNeXt-based generators, they have only been used in GAN frameworks and have limited performance in multi-speaker settings. Moreover, diffusion models, despite training faster than GANs, have slow CPU inference. In this paper, we introduce WaveNeXt 2, a unified ConvNeXt-based framework compatible with both GAN and diffusion vocoders. Its core innovation is residual denoising and sub-modeling, where each sub-model progressively refines the waveform. Experimental results in the multi-speaker dataset demonstrate the effectiveness of our approach: (1) GAN-WaveNeXt 2 is much faster than HiFi-GAN and WaveFit, and (2) Diff-WaveNeXt 2 also delivers much faster inference and competitive synthesis quality compared with FastDiff with 4 steps. The Diff-WaveNeXt 2 is very training-efficient, training in only 32 hours, making it ideal for resource-constrained applications.

While current emotional Text-to-Speech (TTS) models have successfully controlled verbal prosody, they often ignore non-verbal vocalizations (NVs), which are essential for authentic human emotion. Although some non-verbal datasets have recently emerged, they often lack high-quality, fine-grained annotations, which restricts a model's ability to precisely control NV generation. To address this limitation, we propose a novel approach for fine-grained non-verbal expression synthesis. We curate and reprocess female NV utterances from the EARS corpus, develop a new annotation scheme using tags to encode NV types, frequencies, and durations, and build an emotional TTS benchmark to demonstrate its effectiveness. Our evaluation shows that while our NV approach leads to minor trade-offs in perceived naturalness, it significantly improves expressiveness (eMOS 4.20) and emotional recognition accuracy (78.8%). Emotion-specific analysis further reveals that NV cues are highly effective for high-arousal emotions like happy (82.5%) and fear (82.7%), and almost perfectly convey sadness (98.3%).

Passive multi-target tracking (MTT) aims to infer the time-varying kinematic and activity states of an unknown number of sources that emit unknown and possibly nonstationary signals, using only noisy mixtures of these signals observed at sensors. Track-before-detect (TBD) methods improve noise robustness by evaluating multi-target hypotheses directly on raw sensor data, without relying on a preceding detection stage. However, existing TBD likelihoods typically assume that the contribution of each active target to the observation is determined solely by its kinematic state. This assumption does not hold in passive sensing scenarios, where the observed mixtures also depend on unknown and possibly nonstationary source signals. To address this issue, we propose subspace TBD, a passive multi-target TBD method that employs a source-signal-insensitive likelihood derived from the complex spherical Student's $t$ (cST) distribution. Instead of explicitly modeling or estimating the nuisance source signals, the method represents each multi-target hypothesis by the subspace spanned by source steering vectors. The cST likelihood then evaluates how well the normalized multichannel mixtures align with this subspace. We conducted acoustic MTT simulations with two moving speakers in noisy, reverberant environments, comparing the proposed method with a baseline consisting of steered response power with phase transform (SRP-PHAT) followed by a sequential Monte Carlo implementation of the generalized labeled multi-Bernoulli filter (SMC-GLMB). The proposed method achieved lower mean optimal subpattern assignment (OSPA) values in all tested conditions.

This paper presents an innovative approach to enhancing machine learning based communication systems, specifically focusing on multiple-input multiple-output (MIMO) configurations using autoencoders. We optimize the transmitter, receiver, and channel simultaneously under conditions of noise and channel fading, aiming to minimize the bit error rate (BER). By incorporating the Rayleigh fading channel a widely recognized model for wireless channel impairments into the autoencoder framework, we directly train the communication system to handle real world conditions. We introduce a novel optimization process tailored for deep learning-based MIMO communication, and thoroughly analyze the resulting BER performance across various signal to noise ratio (SNR) levels. Our simulation results reveal that the proposed end-to-end wireless communication system achieves significantly lower BER compared to conventional block-based processing methods, highlighting its potential for more efficient and reliable wireless communication.

We examine how aircraft seat configuration interacts with daily operation in Regional Air Mobility by applying a joint supply-demand optimization framework that simultaneously determines market share, fare, and flight schedule. The framework integrates a binary logit discrete choice model into a task assignment formulation, capturing passengers' mode choice between Regional Air Mobility and driving across spatiotemporal origin-destination pairs. We evaluate three U.S. college town corridors under 4-, 6-, and 8-seat configurations across cost scales from 0.4 to 1.0 and fleet sizes from 12 to 30 aircraft. Profitability and throughput serve as primary performance metrics, and we analyze pricing power, operating cost, and revenue to explain performance variation across markets. We find that larger aircraft configurations and fleet sizes do not improve profitability universally. Larger aircraft are preferred where economies of scale are favorable and demand is sufficient and directionally balanced. The best configuration in these case studies is the 4-seat in imbalanced markets and the 6-seat in balanced or dense markets.

The 3GPP V2X resource allocation framework offers two entity classes for coordinating vehicle communication — base-station scheduling and autonomous vehicle selection — a design space we show to be structurally incomplete for safety-critical coordination. We propose Mode 0, a third entity class centered on the Roadside Computing Unit (RCU): infrastructure integrating elevated sensing, sidelink communication, and local computation, owned by traffic management authorities rather than network operators. We formalize Mode 0's subfamily taxonomy (Mode 0a through 0c), its traffic risk taxonomy and mandatory-escalation boundary, its cross-domain coexistence with Modes 1–4, and its standardization pathway. Of the three failure modes motivating this proposal — scheduling saturation, occlusion-driven information gaps, and an institutional authority gap during severe hazards — only scheduling is evaluated by simulation; information and authority are argued architecturally and evidenced by independent deployment convergence across Chinese local and industry-alliance standards, Chinese operator infrastructure, and European and US C-V2X programs, properties no channel-level simulation can instantiate. In a multi-agent reinforcement learning simulation (88 training runs, 95% confidence intervals), per-vehicle policies (Mode 0c) raise worst-interval delivery reliability at the critical safety load from 0.217 under shared per-class policies (Mode 0a) to 0.792, cutting lost safety messages from 38% to 6%. An idealized model of standardized sensing-based semi-persistent scheduling loses 10% at the same load, and per-vehicle coordination more than halves the longest gaps between safety updates against it (3.9 against 8.4 message periods). With spare spectrum, a safety partition sized by the RCU delivers every safety message. We call for a 3GPP study item to formalize Mode 0.

The convergence of large language models (LLMs) with 6G networks is fostering a paradigm of autonomous multi-agent cooperation, which in turn is expected to substantially increase east-west traffic. Although latent-space interaction mechanisms can enable more efficient collaboration than symbolic natural-language (NL) exchanges, prior work often abstracts away the associated communication overhead under practical wireless constraints. In embodied multi-agent settings, heterogeneous interaction media incur disparate inference and transmission costs, thereby inducing an inherent end-to-end (E2E) latency trade-off. To address this, we propose a joint design that integrates communication-media selection with wireless resource allocation. Through analytical characterization and simulation-based evaluation, we show that neither token-based transmission nor key-value (KV) cache-based transmission is uniformly optimal across operating regimes, as performance depends critically on system parameters such as available computational resources and channel conditions. Accordingly, we formulate a joint optimization problem aimed at minimizing the E2E latency of multi-agent collaboration and develop a low-complexity joint media selection and resource allocation (JMSRA) algorithm. Numerical results further confirm that, by adaptively coordinating the interaction media and bandwidth allocation over heterogeneous links, the proposed scheme achieves markedly reduced E2E latency relative to conventional NL-only and KV-cache-only baselines, enabling efficient and robust multi-agent collaboration in future wireless networks.

Cascaded Automatic Speech Recognition - Large Language Model (ASR-LLM) pipelines remain popular for industrial Spoken Dialogue Systems (SDS), primarily because their decoupled design ensures perceptual verifiability. However, cascaded systems suffer from error propagation, as transcription failures inevitably cascade to subsequent components, thereby degrading the final interaction quality. Although ASR confidence scores offer a simple filter for unreliable inputs, this approach is fundamentally limited because it typically fails to detect deletion errors or to distinguish between acoustic (inability to hear clearly) and linguistic (inability to understand) mismatches, both of which require targeted recovery strategies. In this paper, we propose a cause-aware error recovery paradigm that fundamentally rethinks robustness in SDS. Unlike traditional confidence filtering, we introduce a suite of small precision-focused detectors that exploit deep ASR latent representations to disentangle token-level errors into perception, comprehension, and deletion failures. This fine-grained diagnostic intelligence empowers the LLM to orchestrate targeted, multi-turn clarification strategies, effectively transforming ambiguous signals into seamless user interactions. Experimental results validate the precision of our approach, which more than doubles the recall on domain-shift errors (57.96% vs. 23.66%) compared to baselines. Crucially, this diagnostic precision yields up to a 31% reduction in WER and a 19% improvement on the downstream task across diverse accents, distortions, and domains.