Papers

In millimeter wave integrated sensing and communication (ISAC) systems for intelligent transportation, radar and communication share spectrum and hardware in a time division manner. Radar rapidly detects and localizes mobile users (MUs), after which communication proceeds through narrow beams identified by radar. Achieving fine Doppler resolution for MU clutter discrimination requires long coherent processing intervals, reducing communication time and throughput. To address this, we propose a reconfigurable architecture for Doppler estimation realized on a system on chip using hardware software codesign. The architecture supports algorithm level reconfiguration, dynamically switching between low-complexity, high-speed FFT-based coarse estimation and high complexity ESPRIT based fine estimation. We introduce modifications to ESPRIT that achieve 6.7 times faster execution while reducing memory and multiplier usage by 79% and 63%, respectively, compared to state of the art approaches, without compromising accuracy. Additionally, the reconfigurable architecture can switch to lower slow time packets under high SNR conditions, improving latency further by 2 times with no loss in performance.

Spatially structured beams have emerged as a promising technology for enhancing spectrum efficiency (SE) in sixth-generation (6G) networks. However, structured beam schemes based on fixed-position antennas (FPAs) fail to fully exploit the array aperture, thereby limiting their topological reconfigurability and adaptability to diverse communication scenarios. To overcome these limitations, this paper proposes a novel structured beam communication framework exploiting movable antennas (MAs) to achieve reconfigurable array topologies. Specifically, we develop an MA-based geometric modeling framework to construct a variety of practical array topologies, thereby enabling the realization of diverse array configurations utilizing a unified hardware platform. Furthermore, we investigate the joint design of the array topology and the structured beamforming vector to efficiently exploit the array aperture and facilitate the multiplexing of orthogonal spatial modes. Accordingly, we formulate the corresponding beam generation and demodulation schemes and derive the channel gains under varying array topologies. We also propose an alternating optimization algorithm to jointly optimize the array topology configuration, the antenna element positions, and the structured beamforming vector, with the aim of maximizing the system SE. Numerical results demonstrate that the proposed joint design significantly enhances the SE compared to conventional FPA schemes. By synergizing the spatial multiplexing degrees of freedom (DoFs) of structured beams with the mobility DoFs of MAs within 2D planar regions, this work establishes a reconfigurable and practical framework for structured beam-based wireless communications.

The prediction of electromagnetic spectra for MXene-based solar absorbers, where MXenes are a family of two-dimensional transition metal carbides and nitrides, is a computationally intensive task traditionally addressed using full-wave solvers. This study introduces an efficient deep learning framework incorporating transfer learning, multi-channel spectral refinement, and Savitzky-Golay smoothing to accelerate and enhance spectral prediction accuracy. The proposed architecture leverages a pretrained MobileNet version 2 model, fine-tuned to predict 102-point absorption spectra from ($64\times64$) metasurface designs. Additionally, the multi-channel spectral refinement module processes the feature map through multiple convolutional channels, enhancing feature extraction, while Savitzky-Golay smoothing mitigates high-frequency noise. Experimental evaluations demonstrate that the proposed model significantly outperforms baseline convolutional neural network and deformable convolutional neural network models, achieving an average root mean squared error of 0.0227, coefficient of determination ($R^2$) of 0.9563, and peak signal-to-noise ratio of 33.10 decibels. The proposed framework presents a scalable and computationally efficient alternative to conventional solvers, positioning it as a viable candidate for rapid spectral prediction in nanophotonic design workflows.

Aerial base stations mounted on unmanned aerial vehicles (UAVs) support next-generation wireless networks in challenging environments such as urban areas, disaster zones, and remote locations. Further, UAV swarms overcome the challenges of limited battery life and other operational constraints of a single UAV. However, tracking mobile users on the ground by each UAV and the corresponding synchronization between the UAVs is a significant issue that must be addressed before this framework can be deployed in reality. Incorporating additional sensing capabilities to facilitate this additional requirement would introduce significant overhead in terms of hardware, cost, and power to each UAV. Instead, we propose an integrated sensing and communications-enabled swarm UAV system, based on the millimeter-wave IEEE 802.11ad protocol. Further, we show that our proposed system is capable of five-dimensional (5-D) ground target sensing (range, Doppler velocity, azimuth, elevation, and polarization) in an urban environment. Numerical experiments using realistic models demonstrate and validate the performance of 5-D sensing using our proposed 802-11ad-aided UAV system.

Subcode-ensemble decoders improve iterative decoding by running multiple decoders in parallel over carefully chosen subcodes, increasing the likelihood that at least one decoder avoids the dominant trapping structures. Achieving strong diversity gains, however, requires constructing many subcodes that satisfy a linear covering property-yet existing approaches lack a systematic way to scale the ensemble size while preserving this property. This paper introduces hierarchical subcode ensemble decoding (HSCED), a new ensemble decoding framework that expands the number of constituent decoders while still guaranteeing linear covering. The key idea is to recursively generate subcode parity constraints in a hierarchical structure so that coverage is maintained at every level, enabling large ensembles with controlled complexity. To demonstrate its effectiveness, we apply HSCED to belief propagation (BP) decoding of polar codes, where dense parity-check matrices induce severe stopping-set effects that limit conventional BP. Simulations confirm that HSCED delivers significant block-error-rate improvements over standard BP and conventional subcode-ensemble decoding under the same decoding-latency constraint.

Prior works have explored multi-armed bandit (MAB) algorithms for the selection of optimal beams for millimeter-wave (mmW) communications between base station and mobile users. However, when the number of beams is large, the existing MAB algorithms are characterized by long exploration times, resulting in poor overall communication throughput. In this work, we propose augmenting the upper confidence bound (UCB) based MAB with integrated sensing and communication (ISAC) to address this limitation. The premise of the work is that the radar and communication functionalities share the same field-of-view and that communication mobile users are detected by the radar as mobile targets. The radar information is used for significantly reducing the number of candidate beams for the UCB, resulting in an overall reduction in the exploration time. Further, the radar information is used to estimate the realignment time in quasi-stationary scenarios. We have realized the MAB and radar signal processing algorithms on the system on chip (SoC) via hardware-software co-design (HSCD) and fixed-point analysis. We demonstrate the significant gain in execution time using accelerators. The simulations consider complex propagation channels involving direct and multipath, with simple and extended radar targets in the presence of significant static clutter. The resulting experiments show that the proposed ISAC-based MAB achieves a 35% reduction in the overall exploration time and 1.4 factor higher throughput as compared to the conventional MAB that is based only on communications.

Aerial insects can effortlessly navigate dense vegetation, whereas similarly sized aerial robots typically depend on offboard sensors and computation to maintain stable flight. This disparity restricts insect-scale robots to operation within motion capture environments, substantially limiting their applicability to tasks such as search-and-rescue and precision agriculture. In this work, we present a 1.29-gram aerial robot capable of hovering and tracking trajectories with solely onboard sensing and computation. The combination of a sensor suite, estimators, and a low-level controller achieved centimeter-scale positional flight accuracy. Additionally, we developed a hierarchical controller in which a human operator provides high-level commands to direct the robot's motion. In a 30-second flight experiment conducted outside a motion capture system, the robot avoided obstacles and ultimately landed on a sunflower. This level of sensing and computational autonomy represents a significant advancement for the aerial microrobotics community, further opening opportunities to explore onboard planning and power autonomy.

This paper experimentally demonstrates the Koopman-Model Predictive Control (K-MPC) for a real AC-DC converter. The converter is typically modeled with a nonlinear time-variant plant. We introduce a new dynamical approach to lifting measurable dynamics from the plant and constructing a linear time-invariant model that is consistent with control objectives of the converter. We show that the lifting approach, combined with the K-MPC controller, performs well across the full experimental system and outperforms existing control strategies in terms of both steady-state and transient responses.

Audio-Visual Speech Recognition (AVSR) leverages both acoustic and visual cues to improve speech recognition under noisy conditions. A central question is how to design a fusion mechanism that allows the model to effectively exploit visual information when the audio signal is degraded, while maintaining strong performance on clean speech. We propose CoBRA (Cross-modal Bottleneck for Robust AVSR), a bottleneck-based fusion framework that introduces a compact set of learnable tokens to mediate cross-modal exchange. By regulating information flow through these tokens, the audio stream can reliably access essential visual cues even under adverse or out-of-domain noise. Despite limited training data, our model surpasses comparable baselines and remains competitive with large-scale systems through noise-adaptive fusion, demonstrating both efficiency and robustness. Ablation studies highlight that the depth of fusion is the most critical factor, underscoring its importance in designing robust AVSR systems.

This paper investigates the problem of attitude and air velocity estimation for fixed-wing unmanned aerial vehicles (UAVs) using IMU measurements and at least one Pitot tube measurement, with almost global asymptotic stability (AGAS) guarantees. A cascade observer architecture is developed, in which a Riccati/Kalman-type filter estimates the body-fixed frame air velocity and the vehicle's tilt using IMU data as inputs and Pitot measurements as outputs. Under mild excitation conditions, the resulting air velocity and tilt estimation error dynamics are shown to be uniformly observable. The estimated tilt is then combined with magnetometer measurements in a nonlinear observer on SO(3) to recover the full attitude. Rigorous analysis establishes AGAS of the overall cascade structure under the uniform observability (UO) condition. The effectiveness of the proposed approach is demonstrated through validation on real flight data.

Semantic communication (SC) aims to improve communication efficiency by transmitting only task-relevant information, termed semantic features (SFs). Most existing SC frameworks achieve this by optimizing the encoder and decoder while treating the channel between them as fixed. In this paper, we introduce a new perspective by explicitly modeling the entire process from the encoder output to the decoder input as an SF channel, resulting in an encoder-SF channel-decoder pipeline. We observe that the SF channel is configurable through transceiver operations, such as power allocation, thereby providing an additional degree of freedom for improving task performance. Inspired by this, we formulate a joint optimization problem for the encoder, SF channel, and decoder under a mutual information constraint between the transmitted and reconstructed SFs. To provide analytical insight, we derive the optimal SF channel in closed form for an analytically tractable setting. Based on this formulation, we develop a unified semantic feature channel optimization framework applicable to both analog and digital SC systems. To realize the SF channel in general communication systems, we further propose a physical-layer calibration strategy that aligns the actual SF channel with the trained one. Simulation results demonstrate that the proposed framework consistently improves task performance across various communication environments.

This paper studies a challenging scenario in a multiple-input multiple-output (MIMO) integrated sensing and communication (ISAC) system where the locations of the sensing target and the communication user are both unknown and random, while only their probability distribution information is known. In this case, how to fully utilize the spatial resources by designing the transmit beamforming such that both sensing and communication can achieve satisfactory performance statistically is a difficult problem, which motivates the study in this paper. Moreover, we aim to reveal if it is desirable to have similar probability distributions for the target and user locations in terms of the ISAC performance. Firstly, based on only probability distribution information, we establish communication and sensing performance metrics via deriving the expected rate or posterior Cramér-Rao bound (PCRB). Then, we formulate the transmit beamforming optimization problem to minimize the PCRB subject to the expected rate constraint, for which the optimal solution is derived. It is unveiled that the rank of the optimal transmit covariance matrix is upper bounded by the summation of MIMO communication channel matrices for all possible user locations. Furthermore, due to the need to cater to multiple target/user locations, we investigate whether dynamically employing different beamforming designs over different time slots improves the performance. It is proven that using a static beamforming strategy is sufficient for achieving the optimal performance. Numerical results validate our analysis, show that ISAC performance improves as the target/user location distributions become similar, and provide useful insights on the BS-user/-target association strategy.

Image reconstruction and image synthesis are important for handling incomplete multimodal imaging data, but existing methods require various task-specific models, complicating training and deployment workflows. We introduce Any2all, a unified framework that addresses this limitation by formulating these disparate tasks as a single virtual inpainting problem. We train a single, unconditional diffusion model on the complete multimodal data stack. This model is then adapted at inference time to “inpaint” all target modalities from any combination of inputs of available clean images or noisy measurements. We validated Any2all on a PET/MR/CT brain dataset. Our results show that Any2all can achieve excellent performance on both multimodal reconstruction and synthesis tasks, consistently yielding images with competitive distortion-based performance and superior perceptual quality over specialized methods.

Next-generation wireless communications promise transformative technologies such as massive multiple-input multiple-output (MIMO), reconfigurable intelligent surfaces (RIS), integrated sensing and communication (ISAC), and fluid antenna systems (FAS). However, deploying these technologies is hindered by large-scale optimization problems with nonconvex constraints. Conventional Euclidean-space methods rely on approximations or relaxations, which degrade performance and incur substantial computational costs. Riemannian manifold optimization (RMO) offers a powerful alternative that directly operates on the manifold defined by the geometric constraints. This approach inherently satisfies the constraints at every optimization step, thereby avoiding the performance degradation and substantial computational costs. In this paper, we first elaborate on the principles of RMO, including the fundamental concepts, tools, and methods, emphasizing its effectiveness for nonconvex problems. We then introduce its applications in advanced wireless communications, showing how constrained problems are reformulated on their natural manifolds and solved using tailored RMO algorithms. Furthermore, we present a case study on secure beamforming in an FAS-assisted non-orthogonal multiple access (NOMA) system, demonstrating RMO's superiority over conventional methods in terms of both performance and computational efficiency.

Accurate, resource-efficient localization and tracking enables numerous location-aware services in next-generation wireless networks. However, existing machine learning-based methods often require large labeled datasets while overlooking spectrum and energy efficiencies. To fill this gap, we propose LocDreamer, a world model (WM)-based framework for joint target tracking and scheduling of localization anchors. LocDreamer learns a WM that captures the latent representation of the target motion and localization environment, thereby generating synthetic measurements to imagine arbitrary anchor deployments. These measurements enable imagination-driven training of both the tracking model and the reinforcement learning (RL)-based anchor scheduler that activates only the most informative anchors, which significantly reduce energy and signaling costs while preserving high tracking accuracy. Experiments on a real-world indoor dataset demonstrate that LocDreamer substantially improves data efficiency and generalization, outperforming conventional Bayesian filter with random scheduling by 37% in tracking accuracy, and achieving 86% of the accuracy of same model trained directly on real data.

In this letter, a dual-bistatic unmanned aerial vehicles (UAVs) tracking system utilizing downlink Long-Term Evolution (LTE) signals is proposed and demonstrated. Particularly, two LTE base stations (BSs) are exploited as illumination sources. Two passive sensing receivers are deployed at different locations to detect the bistatic Doppler frequencies of the target UAV at different directions according to downlink signals transmitted from their corresponding BSs, such that the velocities of the UAV versus time can be estimated. Hence, the trajectories of the target UAV can be reconstructed. Although both the target UAV and the sensing receivers are around 200 meters away from the illuminating BSs, it is demonstrated by experiments that the tracking errors are below 50 centimeters for 90% of the complicated trajectories, when the distances between the UAV and sensing receivers are less than 30 meters. Note this accuracy is significantly better than the ranging resolution of LTE signals, high-accuracy trajectory tracking for UAV might be feasible via multi-angle bistatic Doppler measurements if the receivers are deployed with a sufficient density.

As sixth generation (6G) standardization accelerates, there is growing consensus in favor of evolutionary waveforms that add new capabilities while preserving compatibility with the orthogonal frequency division multiplexing (OFDM) core of 4G and 5G. This article positions affine frequency division multiplexing (AFDM) as such a candidate, providing structural robustness for high-mobility communications and integrated sensing and communication (ISAC) over doubly dispersive channels while remaining backward-compatible with the legacy OFDM air interface. We first develop a generalized fractional-delay-fractional-Doppler (FDFD) channel model that accounts for practical pulse-shaping filters and the resulting inter-sample coupling. Building on this model, we show that the AFDM transceiver reuses nearly the entire OFDM chain, adding only lightweight digital pre- and post-processing. We then analyze the impact of hardware impairments such as phase noise and carrier frequency offset, and examine the advanced functionalities enabled by the chirp-parameter domain, including index modulation and physical-layer security. Assessing reusability across the radio-frequency, physical, and higher layers, we conclude that AFDM offers an efficient path toward high-fidelity later versions of 6G and beyond (6G+) communications.

Current audio formats present a fundamental trade-off between file size and functionality: lossless formats like FLAC preserve quality but lack adaptability, while lossy formats reduce size at the cost of fidelity and offer no stem-level access.We introduce the Stem-Native Codec (SNC), a novel audio container format that stores music as independently encoded stems plus a low-energy mastering residual. By exploiting the lower information entropy of separated stems compared to mixed audio, SNC achieves a 38.2% file size reduction versus FLAC (7.76 MB vs. 12.55 MB for a 2:18 test track) while maintaining perceptual transparency (STOI = 0.996). Unlike existing formats, SNC enables context-aware adaptive playback, spatial audio rendering, and user-controlled remixing without requiring additional storage. Our experimental validation demonstrates that the stems-plus residual architecture successfully decouples the conflicting requirements of compression efficiency and feature richness, offering a practical path toward next-generation audio distribution systems.

This paper addresses the robust ${\cal H}_2$ synthesis problem for linear fractional transformation (LFT) systems subject to structured uncertainty (parameter) and white-noise disturbances. By introducing an intermediate matrix variable, we derive convex synthesis conditions in terms of linear matrix inequalities (LMIs) that enable both robust and gain-scheduled controller design for parameter-dependent systems. The proposed framework preserves the classical white-noise and impulse-response interpretation of the ${\cal H}_2$ criterion while providing certified robustness guarantees, thereby extending optimal ${\cal H}_2$ control beyond the linear time-invariant setting. Numerical and application examples demonstrate that the resulting robust ${\cal H}_2$ controllers achieve significantly reduced conservatism and improved disturbance rejection compared with conventional robust ${\cal H}_\infty$-based designs.

Predicting interdependent load values in multiport scatterers is challenging due to high dimensionality and complex dependence between impedance and scattering ability, yet this prediction remains crucial for the design of communication and measurement systems. In this paper, we propose a two-stage cluster-then-predict framework for multiple load values prediction task in multiport scatterers. The proposed cluster-then-predict approach effectively captures the underlying functional relation between S-parameters and corresponding load impedances, achieving up to a 46% reduction in Root Mean Square Error (RMSE) compared to the baseline when applied to gradient boosting (GB). This improvement is consistent across various clustering and regression methods. Furthermore, we introduce the Real-world Unified Index (RUI), a metric for quantitative analysis of trade-offs among multiple metrics with conflicting objectives and different scales, suitable for performance assessment in realistic scenarios. Based on RUI, the combination of K-means clustering and k-nearest neighbors (KNN) is identified as the optimal setup for the analyzed multiport scatterer.

Deploying autonomous agents in the wild requires reliable safeguards against tool use failures. We propose a training free guardrail based on spectral analysis of attention topology that complements supervised approaches. On Llama 3.1 8B, our method achieves 97.7% recall with multi-feature detection and 86.1% recall with 81.0% precision for balanced deployment, without requiring any labeled training data. Most remarkably, we discover that single layer spectral features act as near-perfect hallucination detectors: Llama L26 Smoothness achieves 98.2% recall (213/217 hallucinations caught) with a single threshold, and Mistral L3 Entropy achieves 94.7% recall. This suggests hallucination is not merely a wrong token but a thermodynamic state change: the model's attention becomes noise when it errs. Through controlled cross-model evaluation on matched domains ($N=1000$, $T=0.3$, same General domain, hallucination rates 20–22%), we reveal the “Loud Liar” phenomenon: Llama 3.1 8B's failures are spectrally catastrophic and dramatically easier to detect, while Mistral 7B achieves the best discrimination (AUC 0.900). These findings establish spectral analysis as a principled, efficient framework for agent safety.

In the presence of occlusions and measurement noise, geometrically accurate scene reconstructions – which fit the sensor data – can still be physically incorrect. For instance, when estimating the poses and shapes of objects in the scene and importing the resulting estimates into a simulator, small errors might translate to implausible configurations including object interpenetration or unstable equilibrium. This makes it difficult to predict the dynamic behavior of the scene using a digital twin, an important step in simulation-based planning and control of contact-rich behaviors. In this paper, we posit that object pose and shape estimation requires reasoning holistically over the scene (instead of reasoning about each object in isolation), accounting for object interactions and physical plausibility. Towards this goal, our first contribution is Picasso, a physics-constrained reconstruction pipeline that builds multi-object scene reconstructions by considering geometry, non-penetration, and physics. Picasso relies on a fast rejection sampling method that reasons over multi-object interactions, leveraging an inferred object contact graph to guide samples. Second, we propose the Picasso dataset, a collection of 10 contact-rich real-world scenes with ground truth annotations, as well as a metric to quantify physical plausibility, which we open-source as part of our benchmark. Finally, we provide an extensive evaluation of Picasso on our newly introduced dataset and on the YCB-V dataset, and show it largely outperforms the state of the art while providing reconstructions that are both physically plausible and more aligned with human intuition.

We propose a model-agnostic trustworthiness layer that equips any foundation model (FM) for power systems with statistically valid prediction intervals. The layer offers two calibration approaches: (i) stratified conformal prediction (SCP), which partitions residuals by contingency severity and grid element, and (ii) kernel-weighted conformal prediction (KCP), which localizes the calibration to each test scenario via scenario representations, yielding tighter, approximately conditional bounds. Using GridFM as a guiding example, we demonstrate the framework on N-k contingency screening for IEEE 24- and 118-bus systems. The trustworthiness layer ensures that over 90% of all critical violations are captured across N-k levels, minimizing missed detections while maintaining up to 5 times fewer false alarms than DC Power Flow. With negligible computational overhead over the underlying FM, this approach enables reliable large-scale security assessment beyond routine N-1 screening.

Target speaker extraction (TSE) aims to extract the speech of a target speaker from mixtures containing multiple competing speakers. Conventional TSE systems predominantly rely on speaker cues, such as pre-enrolled speech, to identify and isolate the target speaker. However, in many practical scenarios, clean enrollment utterances are unavailable, limiting the applicability of existing approaches. In this work, we propose DAE-TSE, a keyword-guided TSE framework that specifies the target speaker through distinct keywords they utter. By leveraging keywords (i.e., partial transcriptions) as cues, our approach provides a flexible and practical alternative to enrollment-based TSE. DAE-TSE follows the Detect-Attend-Extract (DAE) paradigm: it first detects the presence of the given keywords, then attends to the corresponding speaker based on the keyword content, and finally extracts the target speech. Experimental results demonstrate that DAE-TSE outperforms standard TSE systems that rely on clean enrollment speech. To the best of our knowledge, this is the first study to utilize partial transcription as a cue for specifying the target speaker in TSE, offering a flexible and practical solution for real-world scenarios. Our code and demo page are now publicly available.

The leader-following consensus problem for general linear multi-agent systems over jointly connected switching networks has been a challenging problem and the solvability of the problem has been limited to the class of linear multi-agent systems whose system matrix is marginally stable. This condition is restrictive since it even excludes the most commonly used double-integrator system. This paper presents a breakthrough by demonstrating that leader-following exponential consensus is achievable for general linear multi-agent systems over jointly connected switching networks, even when the system matrix is exponentially unstable. The degree of instability can be explicitly characterized by two key quantities that arise from the jointly connected condition on a switching graph. By exploiting duality, we further show that the output-based distributed observer design problem for a general leader system is solvable over jointly connected switching networks, even when the system matrix is exponentially unstable. This is also in sharp contrast to the existing distributed observers, which rely on the assumption that the leader system is marginally stable.

Ensuring physical-layer security in non-terrestrial networks (NTNs) is challenging due to their global coverage and multi-hop relaying across heterogeneous network layers, where the locations and channels of potential eavesdroppers are typically unknown. In this work, we derive a tractable closedform expression of the end-to-end secure connection probability (SCP) of multi-hop relay routes under heterogeneous Rician fading. The resulting formula shares the same functional form as prior Rayleigh-based approximations but for the coefficients, thereby providing analytical support for the effectiveness of heuristic posterior coefficient calibration adopted in prior work. Numerical experiments under various conditions show that the proposed scheme estimates the SCP with an 1%p error in most cases; and doubles the accuracy compared with the conventional scheme even in the worst case. As a case study, we apply the proposed framework to real-world space-air-groundsea integrated network dataset, showing that the derived SCP accurately captures observed security trends in practical settings.

Large language models (LLMs) offer significant potential for intelligent mobile services but are computationally intensive for resource-constrained devices. Mobile edge computing (MEC) allows such devices to offload inference tasks to edge servers (ESs), yet introduces latency due to communication and serverside queuing, especially in multi-user environments. In this work, we propose an uncertainty-aware offloading framework that dynamically decides whether to perform inference locally or offload it to the ES, based on token-level uncertainty and resource constraints. We define a margin-based token-level uncertainty metric and demonstrate its correlation with model accuracy. Leveraging this metric, we design a greedy offloading algorithm (GOA) that minimizes delay while maintaining accuracy by prioritizing offloading for highuncertainty queries. Our experiments show that GOA consistently achieves a favorable trade-off, outperforming baseline strategies in both accuracy and latency across varying user densities, and operates with practical computation time. These results establish GOA as a scalable and effective solution for LLM inference in MEC environments.

Longer stays at healthcare facilities, driven by uncertain patient load, inefficient patient flow, and lack of real-time information about medical care, pose significant challenges for patients and healthcare providers. Providing patients with estimates of their expected real-time length of stay (RT-LOS), generated as a function of the operational state of the healthcare facility at their anticipated time of arrival (as opposed to estimates of average LOS), can help them make informed decisions regarding which facility to visit within a network. In this study, we develop a healthcare facility assignment (HFA) algorithm that assigns healthcare facilities to patients using RT-LOS predictions at facilities within the network of interest. We describe the generation of RT-LOS predictions via two methodologies: (a) an analytical queuing-theoretic approach, and (b) a hybrid simulation-driven machine learning approach. Because RT-LOS predictors are highly specific to the queuing system in question, we illustrate the development of RT-LOS predictors using both approaches by considering the outpatient experience at primary health centers. Via computational experiments, we compare outcomes from the implementation of the RT-HFA algorithm with both RT-LOS predictors to the case where patients visit the facility of their choice. Computational experiments also indicated that the RT-HFA algorithm substantially reduced patient wait times and LOS at congested facilities and led to more equitable utilization of medical resources at facilities across the network. Finally, we show numerically that the effectiveness of the RT-HFA algorithm in improving outcomes is contingent on the level of compliance with the assignment decision.

Learning the topology of higher-order networks from data is a fundamental challenge in many signal processing and machine learning applications. Simplicial complexes provide a principled framework for modeling multi-way interactions, yet learning their structure is challenging due to the strong coupling across different simplicial levels imposed by the inclusion property. In this work, we propose a joint framework for simplicial complex learning that enforces the inclusion property through a linear constraint, enabling the formulation of the problem as a binary linear program. The objective function consists of a combination of smoothness measures across all considered simplicial levels, allowing for the incorporation of arbitrary smoothness criteria. This formulation enables the simultaneous estimation of edges and higher-order simplices within a single optimization problem. Experiments on simulated and real-world data demonstrate that the proposed joint approach outperforms hierarchical and greedy baselines, while more faithfully enforcing higher-order structural priors.

While high altitude platform stations (HAPS) have been primarily explored as network infrastructure for communication services, their advantageous characteristics also make them promising candidates for augmenting GNSS localization. This paper proposes a metaheuristic framework to jointly optimize the number and placement of HAPS for GNSS enhancement in dense urban environments, considering practical constraints such as elevation masks, altitude limits, and ray-traced visibility from 3D city models. The problem is highly nonconvex due to the discrete HAPS count and the environment-dependent 3D Cramer-Rao lower bound (CRLB). To address this, we develop a tailored version of the adaptive special-crowding distance non-dominated sorting genetic algorithm II (ASDNSGA-II). Simulations show the method successfully identifies the minimum number of HAPS needed to satisfy a CRLB threshold and selects the configuration with the lowest CRLB within that minimum, offering a cost-effective and scalable solution for future HAPS-aided positioning systems.