Papers

We introduce Phantom Twist, a type of single-propeller UAV designed to achieve low visibility through high-speed spinning and the exploitation of motion blur. We develop a two-stage automated design pipeline that optimizes the placement of functional components including batteries, control PCB, motor-propeller assembly, and counterweights. The pipeline minimizes visibility as measured by a human-aligned perceptual metric (LPIPS) while strictly satisfying inertial and aerodynamic constraints required for stable flight. We validate this approach through fabrication and flight testing of multiple prototypes. These tests confirm that our pipeline produces stable, controllable designs and that the optimized UAV exhibits significantly reduced visual perceptibility compared to conventional quadcopters.

Reliable adaptive beamforming is critical for large microphone arrays operating in highly dynamic acoustic environments. In scenarios characterized by fast-moving talkers and interferers, the available sample support for estimating the spatial correlation matrix is often snapshot-deficient. This deficiency degrades the White Noise Gain (WNG), leading to severe target signal cancellation. To ensure stable and robust beamforming, we previously proposed an adaptive diagonal loading method that leverages the Kantorovich inequality to guarantee the WNG remains strictly within specified bounds. However, accurately determining the smallest necessary loading level requires calculating the extreme eigenvalues of the spatial correlation matrix, a computationally expensive $\mathcal{O}(M^3)$ operation for large arrays. In this paper, we introduce a highly efficient $\mathcal{O}(kM^2)$ estimation technique using Lanczos iterations to build a small Krylov subspace. By projecting the correlation matrix onto a tridiagonal matrix of dimension $k \ll M$, we extract Ritz values that rapidly converge to the exact extreme eigenvalues. Our evaluations demonstrate that this Lanczos-accelerated approach achieves performance identical to exact Eigenvalue Decomposition (EVD), ensuring optimal interference suppression and strict WNG adherence at a fraction of the computational cost.

The integration of communication and computation is essential for next-generation wireless systems, especially in scenarios demanding massive connectivity and ultra-low latency. Over-the-air federated learning (AirFL), leveraging the superposition nature of wireless channels, enables fast data aggregation, while non-orthogonal multiple access (NOMA) offers spectrum-efficient connectivity. This paper investigates a fluid antenna (FA)-aided hybrid network, supporting hybrid users comprising both AirFL and NOMA participants. The dynamic reconfigurability of FAs offers significant potential for mitigating interference and enhancing network performance by adapting antenna positions in response to changing channel conditions. We consider practical challenges arising from imperfect channel state information (CSI) and residual interference due to imperfect successive interference cancellation (SIC). To jointly evaluate the learning and communication performance, a hybrid rate metric is introduced. Subsequently, we formulate a robust optimization problem that jointly minimizes the aggregation error while ensuring reliable user communication under CSI and SIC uncertainties. This joint optimization is formulated as a non-convex problem, complicated by the intricate interactions between NOMA and AirFL users and the impact of imperfect CSI and SIC. To solve this problem effectively, we reformulate the optimization as a Markov decision process and solve it using a long short-term memory deep deterministic policy gradient (LSTM-DDPG) algorithm, a memory-based approach within the realm of deep reinforcement learning. Simulation results demonstrate the superiority of the proposed FA-assisted approach over fixed-antenna baselines, particularly under imperfect CSI and SIC conditions, in terms of hybrid rate performance.

This paper presents a compact, low-power, direct RF multi-phase-shift keying (PSK) transmitter (TX) that eliminates the need for a phase-locked loop (PLL) by performing phase modulation directly within a ring oscillator. The proposed architecture exploits synchronized charge extraction at the oscillator's transition points to induce controlled phase shifts while maintaining constant amplitude and frequency. A time-domain multi-triggering technique is introduced to enable reconfigurable multi-mode modulation, supporting 16-PSK, 8-PSK, QPSK, and BPSK within a unified hardware structure. The TX circuit is fabricated in a 22-nm FD-SOI process and operates in the ISM band at 2.4 GHz. Measurement results indicate a symbol rate of 2 MSps with a maximum error vector magnitude (EVM) of 5.13% rms. The core TX occupies 23 {\times} 17.6 {\mu}m2 and consumes 236 {\mu}W, excluding the output driver, which delivers -10 dBm output power over a 60 MHz bandwidth. The proposed design achieves a favorable trade-off between power consumption, circuit complexity, and modulation flexibility, making it well-suited for low-power wireless applications.

Model-Based Systems Engineering (MBSE) is widely treated as the backbone of digital engineering, with languages such as the Systems Modeling Language (SysML) providing the means to capture system structure, behaviour, and verification intent. Yet once verification moves to hardware, the system model is routinely left behind. Domain-specific simulation environments, model transformations, and bespoke tool integrations take over, and the model that began as the authoritative reference drifts out of sync with the implementation it was meant to govern. This paper introduces the SysML Hardware Interface Architecture (SHIA), which keeps an executable SysML model directly inside the verification loop, exchanging messages with physical hardware without intermediate transformation chains, co-simulation platforms, or broker-mediated plugins. SHIA is realised through a SysML side server, written in embedded C++ within IBM Rhapsody, and a hardware side server running on a Raspberry Pi, together establishing a bidirectional link between the digital model and the physical system. A logic gate case study demonstrates the approach end-to-end, from hardware model construction and prototype assembly to test harness design, behavioural statechart control, and staged verification of each component before integration. The integrated system exchanged messages correctly in both directions, and Karnaugh map comparison between the SysML-generated and hardware-generated outputs showed zero discrepancy. The result shows that, when paired with a suitable interface, SysML need not remain a static description that informs downstream tools; it can serve as the executable layer through which hardware behaviour is stimulated, observed, and verified. The work demonstrates a route to model-governed verification and a shorter digital thread between system architecture and the hardware that realises it.

Short-term plasticity (STP) is fundamental to temporal information processing in biological neural systems but remains difficult to realize efficiently in neuromorphic hardware. Memristive electrochemical random-access memory (ECRAM) devices naturally exhibit non-equilibrium ionic dynamics that produce transient conductance modulation; however, these behaviors are typically treated as undesirable variability or tolerated as side effects in memory-centric computing paradigms. In this work, we instead transform these volatile dynamics from a tolerated device artifact into a computational resource through a cross-layer device-circuit-system co-design framework. We introduce a delay-feedback leaky integrate-and-fire (LIF) neuron architecture co-designed with ECRAM synapses that exploits activity-dependent conductance modulation with negligible additional circuit overhead. The architecture integrates ECRAM-based synapses with a tunable delay-feedback spike-generation path, enabling transient device dynamics to directly modulate neuron excitability and synaptic efficacy. We used experimentally characterized ECRAM devices exhibiting transient conductance modulation (1.5 KOhms per spike) to develop a compact behavioral model suitable for circuit-level simulation. Circuit simulations demonstrate two key STP behaviors – synaptic facilitation and intrinsic excitability modulation – while consuming 2 pJ per spike, and the same device-driven mechanisms extend across multiple neuron topologies. Network-level analysis further demonstrates frequency-selective spike processing, allowing individual synapses to act as tunable temporal filters within spiking neural networks. This work demonstrates that non-equilibrium ECRAM dynamics can serve as a native hardware substrate for STP and temporal computation in neuromorphic circuits.

Local interaction laws governing multi-agent systems can be difficult to recover from trajectory data, even when the dynamics are observed faithfully. In systems governed by a nonlinear sheaf Laplacian – a generalization of the graph Laplacian accommodating heterogeneous state spaces and asymmetric communication channels – the coordination law is encoded by edge potential functions whose gradients produce the inter-agent forces. Because trajectory observations record node-state evolution, they expose only the aggregate effect of the edge forces at each node: distinct interaction laws that agree at the node level are indistinguishable from trajectory data alone. We show that the fundamental obstruction to recovery is topological, measured by sheaf cohomology, and that unique recovery from an unconstrained function class is possible if and only if this cohomology vanishes. When the obstruction is nontrivial, we show that recovery within a finite-dimensional parameterized class is possible precisely when a data-dependent information matrix is positive definite. Experiments validate the theory and illustrate that accurate trajectory reproduction need not certify recovery of the underlying interaction law.

Protection system design for multi-terminal HVDC grids is challenging due to the complexity of the system and the often conflicting design requirements. Effective specification of protection component parameters (e.g., DC circuit breakers and series DC inductors) during component-level design is crucial due to interdependencies among components, the need for detailed modeling, and the complex interactions between the protection system and converter control systems. Both analytical and simulation-based approaches have been proposed as solutions for component-level design. However, analytical methods may not accurately represent system behavior given that approximation is necessary, and simulation-based approaches often require extensive computational effort and time. Therefore, this paper presents an efficient systematic design method, combining both approaches. First, a fundamental analytical solution is derived to consider the protection system requirements. Then, a hybrid analytical–EMT methodology is proposed to accelerate convergence toward the required design parameters, after which detailed models are applied to ensure accuracy in design and validation. The approach is applicable to component-level design for both fully and partially selective protection strategies in HVDC grids.

Surface electromyography (sEMG) is a noninvasive technique widely used to control myoelectric prostheses and other human-machine interfaces. However, the high cost of commercial systems limits accessibility in academic and research environments, especially in developing countries. This study presents a validation protocol for a low-cost eight-electrode sEMG wristband prototype based on IEC 60601 and ANSI/AAMI EC13 standards. The protocol includes electrical safety tests, such as leakage current measurement, insulation evaluation, and continuity verification between electrodes and circuits. Functional performance was evaluated by comparing signals acquired with the prototype against those obtained from a commercial reference device (PortiLab2) using Pearson correlation, Bland-Altman analysis, and mean squared error. Additional tests included signal stability during rest and contraction, UART and Bluetooth communication, frequency response, mechanical characterization of the casing, and user comfort assessment. Results showed leakage currents between 11.4 uA and 13.5 uA, adequate insulation, stable signal acquisition, and high correlation with the reference system (r > 0.85). Reliable wireless transmission without packet loss was also observed. Limitations included power supply constraints during wireless testing and discrepancies in the frequency response at high-gain stages compared with simulations. Mechanical tests showed elastic behavior of the casing under loads up to 98 N. The proposed protocol provides a practical and reproducible framework for the technical and functional validation of low-cost sEMG systems for research and educational applications.

Modern power systems increasingly rely on power electronic converters, yet many of these devices are provided as black-box models, limiting the applicability of conventional small-signal analysis (SSA) tools. This work presents a unified multi-variable fitted state-space (SSA-FITSS) methodology that enables accurate small-signal modeling of black-box Voltage Source Converters (VSCs) using frequency-domain (FD) identification, adaptive pole-expansion, and reduced-order realization. The method includes an automated state-interpretation strategy that assigns fitted states to representative control-loop categories based on their dominant frequency ranges, providing an approximate but meaningful physical interpretation of the identified dynamics. This capability allows extensive modal analysis, including eigenvalue sensitivities and participation factors, in systems where internal converter details are unavailable. The methodology is validated on a grid-following (GFL) VSC and applied to the New England system, which contains multiple black-box converters operating in both GFL and grid-forming (GFM) modes. Results show that the SSA-FITSS models accurately reproduce converter and system dynamics, support full eigenvalue-based analysis, and reveal stability limits under varying synchronous generation and GFL penetration levels. The approach overcomes key limitations of existing identification-based techniques by enabling scalable, interpretable, and system-wide stability assessment.

The paper presents a sensitivity analysis of the factors affecting the optimal partitioning of a district heating network for distributed control. Leveraging a physics-based, distributed model predictive control framework and a performance-based partitioning method, this work studies the relationship between variations in system parameters and the resulting optimal partition, providing insight into the robustness of a nominally designed partition to perturbed operating conditions. The enabling methodology is a learning-enhanced branch and bound method that culls the search space, reducing the number of partitions evaluated for each case. The sensitivity of the nominally optimal partition is characterized across twelve parameter variations, including supply temperature, operating season, building flexibility, pipe characteristics, and building type. This simulation study shows that a well-designed nominal partition exhibits an average cost increase of only 2.8% relative to centralized control across eleven of the twelve cases, with three cases identifying the nominal partition as globally optimal under the perturbed conditions. The robustness study is followed by an analysis of the sensitivity of the optimality loss metric (OLM), revealing that, in five of twelve cases, the case-specific OLM-minimizing partitions underperform the nominally optimal one due to shifts in the relative magnitude of heat loss versus flexibility costs. This indicates that proper tuning of cost function weights and initial conditions for the performance optimization problem is essential for reliable partition selection, and that seasonal repartitioning is warranted when demand profiles deviate substantially from the nominal, as observed in the November operating case.

Designing the sensing architecture for large-scale spatio-temporal systems is hard when accuracy requirements are specified but sensor models are uncertain or unavailable. Classical design treats sensor placement and estimation sequentially, requiring valid forward models for each sensing modality. This paper inverts the design flow: given an error budget, synthesize the measurement likelihood that enforces it while injecting minimal information beyond the dynamical prior. The likelihood is constructed by constrained optimization: among all posteriors satisfying a prescribed accuracy bound relative to a target, select the one minimizing Kullback-Leibler divergence from the prior. The solution is a maximum-entropy posterior in relative-entropy form, and the induced likelihood is the Radon-Nikodym derivative. The framework accommodates arbitrary discrepancies and is instantiated for Wasserstein distance, maximum mean discrepancy, $f$-divergences, moment constraints, and hybrid metrics. For each, we derive the discrete particle-level problem, analyze its convex or convex-relaxed structure, and present solvers with complexity scaling. A closed-form solution exists for the symmetric exponential-tilt case, and a distillation procedure converts nonparametric likelihood samples into parametric forms. A two-layer sensor design architecture embeds the synthesized likelihood in the recursive predict-update loop, connecting accuracy budgets to physical sensor placement, precision, and configuration. Numerical experiments comparing four metrics on unimodal and multimodal scenarios confirm the accuracy constraints are reliably enforced and reveal how metric choice determines the amount and spatial distribution of injected information.

In this paper, we present a two-layer architecture for bearing-only sensor placement that improves upon classical D-optimal design. The first layer reweights particles by minimizing Kullback-Leibler divergence from the current distribution subject to a distributional accuracy bound, concentrating mass on regions where the posterior is likely to settle, without reference to the sensor model. The second layer performs D-optimal sensor placement with respect to the reweighted Fisher information matrix, steering sensors toward geometrically informative configurations. Because the two layers are structurally decoupled, the reweighting generalizes across sensing modalities while the placement remains specific to bearing geometry. Systematic experiments on multi-source localization at two noise levels show that this reweighting reduces localization error on average, with the benefit growing as the sensor-to-source ratio increases and as measurements become more informative. The improvement is established in the first few iterations of the sequential procedure and persists as the posterior concentrates.

Neural warm starts can sharply reduce the number of Newton-Raphson iterations required to solve the AC power flow problem, but existing supervised approaches generalize poorly on heavily loaded instances near voltage collapse. We prove a lower bound on the Newton-Raphson iteration count that depends on the direction of the warm start error rather than on its magnitude, and show as a corollary that the bound becomes vacuous as the smallest singular value of the power-flow Jacobian shrinks, identifying the failure mode of supervised regression near the saddle-node bifurcation. Motivated by this analysis, we introduce Newton's Lantern, a finetuning pipeline that combines group relative policy optimization with a learned reward model trained on perturbations of the base model's predictions, using the iteration count itself as the supervisory signal. Across IEEE 118-bus, GOC 500-bus, and GOC 2000-bus benchmarks, Newton's Lantern is the only method that converges on every test snapshot while attaining the smallest mean iteration count.

Today's inference-time workloads increasingly depend on timely access to a model's internal states. We present DMI-Lib, a high-speed deep model inspector that treats internal observability as a first-class systems primitive, decoupling it from the inference hot path via an asynchronous observability substrate built from Ring^2, a GPU-CPU memory abstraction for capturing and staging tensors, and a policy-controlled host backend that exports them. DMI-Lib enables the placement of observation points across a rich space of internal signals and diverse inference backends while preserving serving optimizations and adhering to tight GPU memory budgets. Our experiments demonstrate that DMI-Lib incurs only 0.4%–6.8% overhead in offline batch inference and an average of 6% in moderate online serving, reducing latency overhead by 2x-15x compared to existing baselines with similar observability features. DMI-Lib is open-sourced at https://github.com/ProjectDMX/DMI.

Split Federated Learning (SplitFed) combines federated and split learning to preserve privacy while reducing client-side computation. However, in medical image segmentation, heterogeneous label quality across clients can significantly degrade performance. We propose SplitFed-CL, a co-learning framework where a global teacher guides local students to detect and refine unreliable annotations. Reliable labels supervise training directly, while unreliable labels are corrected via weighted student–teacher refinement. SplitFed-CL further incorporates consistency regularization for robustness to input perturbations and a trainable weighting module to balance loss terms adaptively. We also introduce a novel difficulty guided strategy to simulate human like boundary centric annotation errors, where the degree of perturbation is governed by shape complexity and the associated annotation difficulty. Experiments on two multiclass segmentation datasets with controlled synthetic noise, together with a binary segmentation dataset containing real-world annotation errors, demonstrate that SplitFed-CL consistently outperforms seven state-of-the-art baselines, yielding improved segmentation quality and robustness.

We reformulate the definition of privacy in the private information retrieval (PIR) problem to accommodate flexible privacy requirements. We focus on graph-replicated PIR, with a generalized privacy requirement, instead of requiring all messages to be private from all servers, during retrieval. Towards this, we define a privacy requirement set for each server, which can be an arbitrary subset of all message indices, as long as the stored message indices are in their privacy requirement set. Since both the storage and privacy requirement sets have many possibilities, we focus on two specific storage settings, namely the path and cyclic graphs. We consider several privacy settings for each of them, which are not necessarily the same, to give different examples for privacy sets. Of particular interest are the privacy sets that comprise the indices of messages stored at servers within a neighborhood range. The neighborhood range parameter allows a transition from the recently introduced local PIR [1] to the standard graph-replicated PIR. In these cases, we derive bounds on the capacity or find the exact capacity.

We rethink the definition of privacy in multi-server, graph-replicated private information retrieval (PIR) systems, and introduce a novel setting where the user's privacy is governed by the servers' storage structure. In particular, while retrieving a message from a server, the user is concerned with hiding their desired message index from the server, only if the server stores the corresponding message. We coin this privacy requirement as local user privacy and the resulting PIR problem as local PIR on the graph. Our goal is to measure the gain in communication efficiency of local PIR, compared to that of canonical PIR, by establishing its capacity, i.e., the maximum number of message symbols retrieved, per downloaded symbol. To this end, we observe a remarkable gain in the local PIR capacity of graphs, that are disjoint union of distinct graphs, which is multiplicative, compared to the PIR capacity, when the individual graphs are identical. For connected graphs, we propose schemes to establish capacity lower bounds for edge-transitive and bipartite graphs, which are greater than the best-known PIR capacity bounds. Finally, we derive the exact local PIR capacity for the cyclic graph, and the path graph with an odd number of vertices.

Cyber-physical system (CPS) forecasting models depend on sensor streams with noisy, biased, missing, or temporally misaligned readings, yet standard forecasting evaluation often selects models by nominal error without showing whether they remain robust under such faults. We introduce SensorFault-Bench, a shared CPS-grounded sensor-fault stress-test protocol for evaluating forecasting architectures and robustness-improvement methods, and an operational taxonomy organizing the method comparison. Across four real-world datasets and eight scored scenarios governed by a standardized severity model, it reports worst-scenario degradation, clean mean squared error (MSE), and worst-scenario fault-time MSE, separating relative robustness from absolute error. A disjoint fault-transfer split lets explicit fault-training methods train on adjacent fault families while evaluation uses separate benchmark scenarios. Empirically, forecasting architectures favored by clean MSE can degrade sharply under faults, and clean-MSE rankings can disagree with worst-scenario fault-time error rankings. Chronos-2, the evaluated zero-shot foundation-model representative, matches or trails the last-value naive forecaster in clean MSE on the two single-target datasets and has the largest worst-scenario degradation on ETTh1 and Traffic, where all channels are forecast targets. For the evaluated robustness-improvement method set, paired deltas show selective degradation reductions: projected gradient descent adversarial training and randomized training lead where value faults dominate observed degradation, while fault augmentation leads where availability faults dominate. SensorFault-Bench provides open-source code, documented data access, and reproduction and extension guides, so new datasets, architectures, and robustness-improvement methods can be evaluated under the same CPS sensor-fault robustness protocol.

Audio-visual large language models (AVLLMs) have recently emerged as a powerful architecture capable of jointly reasoning over audio, visual, and textual modalities. In AVLLMs, the bidirectional interaction between audio and video modalities introduces intricate processing dynamics, necessitating a deeper understanding of their internal mechanisms. However, unlike extensively studied text-only or large vision language models, the internal workings of AVLLMs remain largely unexplored. In this paper, we focus on cross-modal information flow between audio and visual modalities in AVLLMs, investigating where information derived from one modality is encoded within the token representations of the other modality. Through an analysis of multiple recent AVLLMs, we uncover two common findings. First, AVLLMs primarily encode integrated audio-visual information in sink tokens. Second, sink tokens do not uniformly hold cross-modal information. Instead, a distinct subset of sink tokens, which we term cross-modal sink tokens, specializes in storing such information. Based on these findings, we further propose a simple training-free hallucination mitigation method by encouraging reliance on integrated cross-modal information within cross-modal sink tokens. Our code is available at https://github.com/kaistmm/crossmodal-hub.

Large language-vision models (LLVM), such as OpenAI's ChatGPT and GPT-4, have gained prominence as powerful tools for analyzing text and imagery. The merging of these data domains represents a significant paradigm shift with far-reaching implications for automatic target recognition (ATR). Recent transformer-based LLVM research has shown substantial improvements for geospatial perception tasks. Our study examines the application of LLVM to remote sensing image captioning and visual question-answering (VQA), with a specific focus on synthetic aperture radar (SAR) imagery. We examine newly published LLVM methods, including CLIP and LLaVA neural network transformer architectures. We have developed a work-in-progress SAR training and evaluation benchmark derived from the MSTAR Public Dataset. This has been extended to include descriptive text captions and question-answer pairs for VQA tasks. This challenge dataset is designed to push the boundaries of an LLVM in identifying nuanced ATR details in SAR imagery. Utilizing parameter-efficient fine-tuning, we train an LLVM method to identify fine-grained target qualities at 98% accuracy. We detail our data setup and experiments, addressing potential pitfalls that could lead to misleading conclusions. Accurately identifying and differentiating military vehicle types in SAR data poses a critical challenge, especially under complex environmental conditions. Mastering this target recognition skill may require a human analyst months of training and years of practice. This research represents a unique effort to apply LLVM to SAR applications, advancing machine-assisted remote sensing ATR for military and intelligence contexts.

This thesis develops a theoretical framework to evaluate the monitoring capability of IoBNT networks. We consider a scenario in which nanosensors passively flow in the bloodstream and detect biomarkers associated with potential diseases, reporting their detections to external gateways on the skin that host a monitoring device. The nanosensors thus realize an artificial point-to-point communication channel between the disease region and the monitor: some packets reach the destination directly, while others are lost through vessel paths that bypass the gateway. We evaluate the network's monitoring capability over this artificial channel using the \ac{AoI} concept, which jointly integrates sample generation (at the disease region), carrying (nanosensor travel through vessels), and delivery (nanosensor-to-gateway) as random events. These are modeled through (i) a Markov model that follows cardiovascular physiology and (ii) channel models of reported nanocommunication technologies. We compute the Markov transition probabilities using a cardiovascular simulator built as a low-complexity electric circuit model of the human vessels. For the nanosensor-to-gateway link, we model two well-known schemes: ultrasonic and terahertz channels. Integrating these components within the \ac{AoI} framework, we report information freshness via the average \ac{PAoI} metric. Under realistic physiological and communication assumptions, fresh information appears on the monitor within tens of seconds. The network is therefore suitable for monitoring tissue-level processes such as bacterial infections, while more adequate architectures are needed to monitor cellular-scale processes, which occur on timescales below tens of seconds.

We introduce SMART-HC-VQA, a Sentinel-2-based visual question answering dataset derived from the IARPA SMART Heavy Construction dataset, designed for spatiotemporal analysis of human activity. The dataset transforms construction-site annotations, construction-type labels, temporal-phase labels, geographic metadata, and observation relationships into natural language question-answer triplets. This approach redefines the existing dataset as a temporally extended automatic target recognition and visual question answering (VQA) challenge, considering a fixed geospatial site as a target whose attributes and activity states evolve across sparse satellite observations. Currently, SMART-HC-VQA comprises 21,837 accessible Sentinel-2 image chips, 65,511 single-image VQA examples, and approximately 2.3 million two-image temporal comparison examples generated via our novel Image-Pairwise Combinatorial Augmentation. We detail the workflow for retrieving and processing Sentinel-2 imagery, segmenting large satellite tiles into site-centered images, maintaining traceability to SMART-HC annotations, and analyzing the distributions of site size, observation count, temporal coverage, construction type, and phase labels. Additionally, we describe an implemented multi-image MLLM training framework based on LLaVA-NeXT Mistral-7B, adapted to accept multiple dated image inputs and train on metadata-derived VQA examples. This work offers a reproducible foundation for understanding language-guided remote sensing activities, aiming not only to detect change but also to reason about the ongoing processes, their progression, and potential future developments.

Decentralized collision avoidance remains challenging, particularly when agents do not communicate any information related to planned trajectories. Most existing approaches either rely on conservative coordination mechanisms or provide limited guarantees on recursive feasibility and convergence. This paper develops a decentralized contingency MPC framework for multi-agent systems with nonlinear dynamics that achieves collision-free motion under a state-only information pattern. Each agent follows the same consensual rule set, enabling safe decentralized planning without communication. Each agent solves a local optimization problem that couples a nominal trajectory with a contingency certificate ensuring a feasible backup maneuver under receding-horizon operation. A novel geometric and decentralized safe-set update mechanism prevents feasibility loss between consecutive time steps. The resulting scheme guarantees recursive feasibility, including collision avoidance, and establishes a Lyapunov-type convergence result to an admissible safe equilibrium. Simulation results demonstrate performance in both sparse and dense multi-agent environments, including cluttered bottleneck scenarios and under plug-and-play operation.

Grey failures in the computing continuum produce ambiguous overlapping symptoms that existing approaches fail to diagnose reliably, either due to a lack of causal awareness or acting under high epistemic uncertainty, risking destructive interventions. This paper presents an uncertainty-aware resilience micro-agent for causal observability (AURORA), a lightweight framework for diagnosing and mitigating grey failures in edge-tier environments. The framework employs parallel micro-agents that integrate the free-energy principle, causal do-calculus, and localized causal state-graphs to support counterfactual root-cause analysis within each fault's Markov blanket. Restricting inference to causally relevant variables reduces computational overhead while preserving diagnostic fidelity. AURORA further introduces a dual-gated execution mechanism that authorizes remediation only when causal confidence is high and predicted epistemic uncertainty is bounded; otherwise, it abstains from local intervention and escalates the diagnostic payload to the fog tier. Our experiments demonstrate that AURORA outperforms baselines, achieving a 0% destructive action rate, while maintaining 62.0% repair accuracy and a 3ms mean time to repair.

This paper introduces a proactive Unmanned Aerial Vehicle (UAV) mobility management xApp for Open Radio Access Network (O-RAN) Near Real-Time Radio Intelligent Controller (Near-RT RIC) environments, employing Double Deep Q-Network (DDQN) reinforcement learning (RL) enhanced with transfer learning to optimise handover decisions for UAVs operating along predetermined flight trajectories. Unlike reactive approaches that respond to signal degradation, the proposed framework anticipates network conditions and minimises both outage probability and handover frequency through predictive optimisation. The system leverages centralised weight averaging to consolidate knowledge from multiple flight scenarios into a global model capable of generalising to previously unseen operational environments without extensive retraining. A comprehensive evaluation demonstrates that the proposed framework achieves a favourable trade-off between handover frequency and connectivity reliability, reducing handover events by up to 54.6% compared to greedy approaches while maintaining outage probability at practically negligible levels. The results validate the effectiveness of intelligent learning-based approaches for UAV mobility management in next-generation O-RAN architectures, thereby contributing to seamless integration of aerial user equipment into cellular networks.

Event identification in continuous neural recordings is a critical task in neuroscience. Decoding in EEG is dominated by classifying windows aligned to known event onsets. However, while available in controlled experiments, such onsets are absent in continuous real-world monitoring. Here, we introduce DANCE, a deep learning pipeline that frames neural decoding as a set-prediction problem and jointly detects and classifies events directly from raw, unaligned signals. Evaluated separately on ten datasets curated from the literature with a wide variety of event types (ranging from milliseconds to minutes in duration), our model outperforms existing methods on a broad range of cognitive, clinical and BCI tasks. This single architecture establishes a new state of the art in the competitive task of seizure monitoring and matches the accuracy of onset-informed models for BCI tasks. Overall, our method marks a step towards end-to-end asynchronous neural decoding models

This paper proposes the joint design of reconfigurable intelligent surfaces (RIS) and zero-forcing (ZF) precoding for the downlink (DL) multiuser multiple-input single-output (MU-MISO) setup in millimeter-wave (mmWave) bands, where ZF is particularly attractive due to its ability to suppress inter-user interference by exploiting the large antenna arrays and sparse directional channels characteristic of mmWave systems. This ensures efficient spatial multiplexing with manageable complexity, making ZF a practical and in modern 5G/6G deployments. However, a careful design is necessary to overcome potential rank deficiency in the channel matrix. For the MU-MISO case, rank deficiency may arise if users exhibit significantly different channel gains or if, being in far-field, they are aligned with the position of the transmitter. On the other hand, the deployment of a RIS introduces artificial scattering which can shape the radio environment to address those situations. We explore the joint design under perfect channel knowledge, assess the impact of imperfect channel estimation on the bit error rate (BER) and propose a robust design of pilot transmissions that equalizes multiuser interference across users in the presence of channel errors in the precoder design. This evaluation shows the advantages of optimized RIS-aided ZF MU-MISO communication for the DL of wireless systems.

Open Radio Access Networks (O-RAN) are increasingly adopting data-driven control through Deep Reinforcement Learning (DRL) to optimize complex tasks such as network slicing and mobility management. However, the deployment of DRL in carrier-grade networks is hindered by its inherent opacity and stochastic execution, which limit operator trust, auditability, and safe deployment. Existing explainable AI (XAI) approaches primarily provide post-hoc insights and fail to produce executable, interpretable policies suitable for operational environments. In this paper, we present DeRAN, a neuro-symbolic framework that bridges the gap between DRL performance and operational transparency by distilling black-box DRL policies into human-readable symbolic representations. DeRAN introduces a concept-driven abstraction layer that transforms high-dimensional network telemetry into a compact set of semantically meaningful features, enabling interpretable policy learning. Building on the semantically grounded concepts, DeRAN synthesizes symbolic policies using deep symbolic regression (DSR) for continuous control and neurally guided differentiable logic (NUDGE) for discrete decision-making. We implement DeRAN on a live 5G O-RAN testbed and evaluate it on two representative use cases. Experimental results demonstrate that DeRAN achieves 78% and 87% of DRL's cumulative rewards in the two use cases, while offering interpretability and auditability by design. Source code is available at https://github.com/Jadejavu/DeRAN

Reachability analysis of neural networks, which seeks to compute or bound the set of outputs attainable over a given input domain, is central to certifying safety and robustness in learning-enabled physical systems. Since exact reachable set computation is generally intractable, existing methods typically rely on tractable overapproximations. Examining the state of the art for smooth, twice-differentiable networks, we observe that existing approaches exploit at most second-order information and do not systematically leverage higher-order information. In this work, we introduce HiTaB, a novel verification framework that exploits second-order smoothness through both the Hessian, $\nabla^2 f$, and its Lipschitz constant, $L_{\nabla^2 f}$. We further develop a unified hierarchy of zeroth-, first-, and second-order bounds, together with precise conditions under which higher-order approximations yield provable improvements. Our main technical contribution is a compositional procedure for efficiently bounding $L_{\nabla^2 f}$ in deep neural networks via layerwise propagation of curvature bounds. We extend the framework to both $\ell_2$- and $\ell_\infty$-constrained input sets and show how it can be integrated into branch-and-bound verification pipelines. To our knowledge, this is the first practical reachability analysis framework for smooth neural networks that systematically exploits Lipschitz continuity of curvature, leading to tighter and more informative safety certificates.