Papers

This paper presents a state observer design for continuous linear time-invariant (LTI) systems subject to unknown bounded disturbances that enforces a prescribed bound on the observer output error (residual). The proposed observer augments a Luenberger observer with state resets triggered when the residual norm reaches a prescribed bound. The reset map guarantees contraction of the residual at jump instants, forward invariance of the prescribed residual envelope, and strict decrease of the estimation-error Lyapunov function at every reset, while preserving the uniform boundedness properties of a standard Luenberger observer. Notably, the observer design does not require knowledge of the disturbance bound. Simulation results confirm the analysis: the residual remains within the prescribed bound, whereas a standard Luenberger observer with the same gains violates this bound.

Hearing aids impose strict latency and power constraints that current DNN-based speech enhancement systems struggle to meet on embedded hardware. We characterize this gap by deploying both speech separation and denoising using the lightweight SuDoRM-RF++ architecture on the AMD-Xilinx Kria KV260, evaluated at FP32 and 16-bit fixed-point precision for each task. Across these configurations, first-sample latency tracks with on-chip parameter caching rather than arithmetic throughput, identifying data movement as the primary bottleneck. Precision reduction halves the model memory footprint without compromising objective speech quality. The fixed-point denoising accelerator achieves a first-sample latency of 9.7~ms, meeting the 10~ms clinical threshold, while speech separation reaches 16.0~ms. These measurements establish concrete resource requirements for embedded DNN-based speech enhancement and quantify the remaining gap to hearing aid deployment.

Randomized smoothing (RS) certifies robustness in the vector space where Gaussian noise is added. In audio classification, this space is often not uniquely defined as standard pipelines normalize, range-control, and transform waveforms into log-mel or other spectral features. We show that direct RS is therefore under-specified unless the certified object and preprocessing policy are explicit. On two audio benchmarks, keyword spotting and environmental-sound classification, we study waveform, feature-space, and post-processed smoothing. Our diagnostics show why representation-aware reporting is necessary: at the same smoothing level $\sigma=0.0025$, the two datasets share the same median raw radius $.007996$, but different waveform energies yield different SNR-equivalent scales ($83.98$ vs. $90.97$ dB); log-mel smoothing gives higher positive-radius certified accuracy on environmental sounds ($68.42\%$ vs. $65.53\%$), certifying more examples with nonzero radius but over features rather than waveforms; and clipping or peak normalization changes the effective perturbation norm by roughly $230$–$351\times$. We therefore recommend that audio RS studies choose and report the task-specific certified object and perturbation model, including the perturbation location, gain policy, raw radius, and any post-noise geometry changes.

This contribution presents an experimental integrated real-time 8 x 8 distributed MIMO (D-MIMO) testbed for wideband backscatter communication (BSC) and wireless power transfer (WPT). The testbed operates in the 2.45 GHz band with coherent sampling at 200 MS/s, employs a backscatter link frequency of 40 kHz, and uses wideband 5G NR reference signals for excitation. We evaluate the testbed by exploiting the estimated channel state information (CSI) in two target applications: wireless power transfer towards the backscatter device (BD) and real-time positioning of a BD in an indoor environment. In conjunction with the baseband processing chain introduced, the testbed requires less than 2 ms of total airtime to excite the system and acquire the signals for subsequent synchronization and CSI estimation on uplink BSC signals. With the CSI, we demonstrate effective energy harvesting gains of up to 12 dB.

Computed Axial Lithography is a volumetric additive manufacturing method that selectively cures photosensitive resin through the 3D superposition of patterns of light, offering advantages over layer-based processes including rapid print times, reduced layer artifacts, and compatibility with high-viscosity materials. However, diffusive effects, primarily those of free-radical quenchers such as oxygen, blur the boundary between cured and uncured regions, limiting resolution and preventing the reproduction of sharp, high-spatial-frequency features. By comparing micro-CT data to computational dose models convolved with kernels across a range of diffusivities, we establish a framework for extracting a single diffusion kernel from any standard uncorrected print to account for all observed deviations from the target. In this work, we correct diffusion-induced blurring by co-optimizing for its effects alongside the inherent blur of the computed tomography reconstruction, demonstrating improved fidelity over previous approaches of pre-compensating the target geometry via deconvolution.

Moving horizon estimation is an efficient technique to estimate states and parameters of constrained dynamical systems. It relies on the solution of a finite horizon optimization problem to compute the estimates, providing a natural framework to handle bounds and constraints on estimates, noises and parameters. However, the approximation of the arrival cost and its updating mechanism are an active research topic. The arrival cost is very important because it provides a mean to incorporate information from previous measurements to the current estimates and it is difficult to estimate its true value. In this work, we exploit the features of adaptive estimation methods to update the parameters of the arrival cost. We show that, having a better approximation of the arrival cost, the size of the optimization problem can be significantly reduced guaranteeing the stability and convergence of the estimates. These properties are illustrated through simulation studies.

Reconfigurable Intelligent Surface (RIS) designs with 1-bit phase resolution often suffer from strong quantization lobes in the far field, which significantly degrade wireless communication performance. This work investigates the minimum phase resolution required for RIS to eliminate far-field quantization lobe. The analysis demonstrates that 2-bit phase discretization offers an optimal balance between performance and hardware complexity. A practical 2-bit RIS unit cell is designed, and a 20 x 20 array configuration is implemented to evaluate its performance. The quantization-lobe suppression capability is validated through full-wave radar cross-section (RCS) simulations under plane-wave illumination for the entire RIS array. The fabricated prototype is further characterized experimentally, achieving a -13.1 dB quantization-lobe level compared to -0.8 dB for its 1-bit counterpart, confirming both the analytical and full-wave simulation results.

Vision-based UAV navigation becomes challenging when navigation targets are distributed across complementary camera views and cannot be reliably observed from a single viewpoint. We propose AgenticDiffusion, an agentic multi-view UAV navigation framework that semantically coordinates first-person-view (FPV) and top-view observations for mission-level navigation. Given a natural-language instruction, AgenticDiffusion identifies the requested targets, selects the most appropriate camera view for each navigation task, determines the corresponding navigation goal, and invokes the appropriate view-conditioned diffusion planner for trajectory generation. The resulting trajectories are executed using Nonlinear Model Predictive Control (NMPC). AgenticDiffusion was evaluated in four real-world indoor scenarios, achieving an overall mission success rate of 80% across 40 physical-flight trials. In mixed-visibility scenarios, where the requested targets were distributed across FPV and top-view observations, coordinated multi-view navigation reduced average mission time by 50.8% relative to FPV-only navigation and by 26.8% relative to Top-only navigation. The semantic view-selection mechanism was also robust to lexical variation in target descriptions, achieving 100% accuracy across 66 test cases, compared with 63.64% for a confidence-based view-selection baseline. In a substantially larger Gazebo environment, AgenticDiffusion achieved a 90% mission success rate and completed the multi-stage mission, whereas the FPV-only and Top-only variants were unable to complete all requested navigation tasks.

Conventional hearing aids rely on fixed, frequency-dependent amplification and compression to manage reduced sensitivity, which often fails to provide sufficient listening support in complex environments, such as situations with multiple speakers (the “cocktail party” problem). To more comprehensively address the underlying encoding dysfunctions of hearing loss, we introduce the Differentiable Auditory Loop (DAL), a new open-source framework for personalized hearing aid design and fitting. Our first implementation of DAL incorporates CARFAC, a differentiable model of human cochlear function, which we ported to JAX, to optimize a deep neural network to match impaired auditory neural activity patterns with a normal-hearing reference. To build a hearing aid with the fine-grained spectro-temporal signal processing required, we adopt SEANet, a waveform-to-waveform fully convolutional UNet generator. We fine-tune the network by comparing the outputs of a CARFAC model fitted to normal hearing with that of a CARFAC model fitted to match each subject's individual hearing impairment. The comparison is done using loss functions derived from the respective CARFAC neural activity pattern (NAP) outputs and stabilized auditory images (SAIs), the latter providing a 2D representation that captures phase-insensitive temporal structure in the auditory nerve output. Through gradient descent, the SEANet model learns to both denoise the input and compensate for the hearing loss modelled by the impaired CARFAC model. Across neural-representation and signal-fidelity metrics, the DAL-optimized SEANet model outperformed the tested master hearing aid (MHA) baselines. The DAL framework provides a practical path toward model-based, machine-learning-driven personalization of hearing aid signal processing. Next steps include hardware deployment to enable real-world clinical testing.

We present a theoretical model for a two-element dynamic phased array and characterize the transfer of information as a function of angle. The array is based on a two-state switched structure with phase shifting to support beamsteering. Dynamic motion of the phase center of antenna arrays generates time-varying radiation patterns that, when appropriately designed, support directional modulation, or the transfer of information to regions of space that are narrower than that covered by the energy radiated by the array. We evaluate the impact of switching frequency and steering on the spatial width of the information beam, which is the region of space where information is recoverable. The concepts are evaluated through simulation and experiment using a 0.75$\lambda$ two-element array operating at 2.5 GHz.

Extreme data scarcity and inherent multipath spatial ambiguity severely limit existing deep learning-based channel state information (CSI) fingerprinting localization schemes for target unmanned aerial vehicles (UAVs). To overcome these challenges, we propose an end-to-end semi-supervised generative localization framework. First, by exploiting the temporal correlations inherent in continuous flight trajectories, a self-supervised encoder extracts robust spatial features from massive unlabeled CSI sequences to establish structured latent representations. Following this, we utilize a consistency model, a powerful derivative of diffusion architectures, as the core generative backbone to map the learned latent space to physical coordinates, jointly fine-tuning the pre-trained encoder with a strictly limited set of labeled CSI. This consistency formulation models the conditional distribution to resolve the mean collapse problem of discriminative models, while compressing the inference trajectory to 1-2 steps to avoid the latency bottleneck of traditional diffusion models. Furthermore, a lightweight distributed fusion mechanism is designed to aggregate spatial predictions across multiple base stations (BS) from a multi-view geometry perspective. Comprehensive evaluations on a real-world measurement dataset demonstrate that our framework achieves low latency and suppresses the mean localization error to 9.77 cm under a 3-BS fusion setup with only a 1% label fraction, significantly outperforming existing fully supervised and semi-supervised discriminative baselines.

Deep learning models are increasingly central to autonomous vehicle (AV) pipelines, yet their integration has traditionally followed a monolithic design where perception, planning, and control execute on a single onboard computer. This design overlooks the emerging paradigm of cooperative autonomy, where vehicles interact with roadside units (RSUs), edge servers, and cloud-hosted intelligence through vehicle-to-everything (V2X) connectivity. Cooperative perception and control improve safety and efficiency, but also introduce systems-level challenges: network latency, compute heterogeneity, and multi-tenant contention, all critically affect real-time decision-making. These challenges are further amplified by the increasing reliance on large foundation models, whose scale necessitates cloud deployment. We present CADET (Cooperative Autonomy through Distributed Experimentation Toolkit), a modular platform for systematic and reproducible evaluation of distributed cooperative autonomy systems under realistic deployment conditions. CADET decouples the AV stack into composable modules that can be flexibly deployed across vehicles, infrastructure, and edge/cloud tiers. The framework integrates state-of-the-art models, incorporates trace-driven network and workload emulation, and provides synchronized model-, system-, and task-level instrumentation. Through V2V and V2I experiments, we show that distributed deployment choices fundamentally shape safety, with V2V intent packets outperforming cloud-based perception and RSU-assisted perception sustaining safety until overloaded by concurrent requests. Although designed for AV pipelines, CADET also supports dataset-driven experimentation, enabling systems and ML researchers to benchmark distributed inference workloads independently of full vehicle simulation. CADET is open source, with code and demo available at https://nesl.github.io/cadet-web.

Reconstructing naturalistic music from EEG requires extracting music-specific information from weak signals distributed across electrodes. Existing spectrogram regression and diffusion approaches provide audio synthesis mechanisms, but learning the EEG-to-music mapping faces an information–estimation tradeoff. Combining electrode measurements into fewer features can ease estimation from limited recordings while obscuring distinctions between music segments. We propose channel-preserving representation alignment, separating electrode information retention from prediction regularization. Per-electrode tokenization preserves spatial identity for learned aggregation, while multi-view self-distillation and channel dropout encourage consistency across electrode subsets during pretraining and music alignment. The aligned representation conditions a pretrained audio generator. Our theory explains these complementary choices. A finite-sample regression analysis identifies when retained task information outweighs estimation cost. For nonlinear alignment, an exact decomposition reveals a prediction-consistency penalty in the masked contrastive objective, with margin-dependent bounds connecting prediction variability to ranking errors across electrode subsets. On NMED-T/H, our method improves within-subject 50-way identification from 0.388 to 0.485 and generated-audio CLAP similarity from 0.635 to 0.683 compared with previous work.

Integrated sensing and communication (ISAC) enables simultaneous sensing and data transmission but exposes a critical vulnerability: probing signals may be intercepted, revealing both the transmitted information and the act of sensing itself. Existing physical layer security approaches mitigate interception yet operate with detectable signals, leaving sensing activity observable to a passive warden. This paper introduces sub-noise-floor pseudo-random probing (SNF-PRP), a covert sensing framework for OFDM-based ISAC systems under an energy-detection adversary model. SNF-PRP establishes an $\epsilon$-covertness guarantee via Kullback-Leibler (KL) divergence, exploits an $N_{\mathrm{sc}}$-fold spreading gain absent from prior wideband analyses, and derives in closed form the minimum integration length required to achieve a target Cramér-Rao bound (CRB). Simulations under 5G~NR n78 numerology confirm sub-0.5\,m range and sub-0.5\,m/s velocity accuracy with KL divergence $5.8\times$ below the covertness threshold, validating joint feasibility at $-12$\,dB and $-15$\,dB probing powers.

Conversational ASR for lower-resource languages and niche domains is limited by the scarcity of domain-matched multi-speaker training data. We propose an augmentation pipeline that generates scenario-level dialogues with participant metadata, maps speaker attributes to TTS voice profiles, and assembles synthesized utterances into speaker-aware simulated conversations. We evaluated five LLM families under single-generator, fixed-budget mixture, and scale-up settings using the same FastConformer-Large training recipe for each one. We ran comprehensive evaluations on the Hungarian BEA-Dialogue benchmark corpus, with the method itself being applicable to any language given the resources for each component. The results show that synthetic conversations consistently improve speech recognition performance, but generator choice and data composition strongly affect the gains. Our largest training configuration, using only 67 hours of real conversations and 636 hours of simulated data, achieves better performance on the evaluation benchmark than a zero-shot model trained on 2700 hours of Hungarian speech. These findings indicate that LLM-generated conversational data synthesized with TTS is a practical complement to real conversational corpora for speech model training.

Motivated by recent developments in stochastic modeling of clock jitter in Analog-to-Digital Converters (ADCs) as autoregressive processes of order one (AR(1)), we study the density and stability properties of AR(1)-jittered sampling sets for Paley-Wiener signals. While the stability of sampling sets under i.i.d. perturbations of the integer lattice is classical, no analogous analysis was previously available for the correlated jitter observed in practice. We show that, despite having the correct asymptotic density both on average and almost surely, such sets almost surely fail to be stable sampling sets. We complement this negative result with a finite-dimensional analysis, providing explicit non-asymptotic conditions on the number of samples $ 2N+1 $, the jitter standard deviation $ \sigma_\xi $ and the confidence level $ 1 - \eta $ under which the smallest singular value of the $ (2N+1) \times (2N+1) $ jittered sinc matrix exceeds a prescribed threshold $ \epsilon $ with probability at least $ 1-\eta $, so that the matrix is well-conditioned with high probability.

In robotics systems, vast amounts of visual data are easily captured at high resolution using low-cost, low-power hardware. Yet, limited bandwidth and on-device compute resources prevent full utilization when transmitted via conventional codecs like JPEG/MPEG. Newer codecs, like AV1/AVIF, improve the rate-distortion trade-off, but demand far more resources for encoding, impractical without custom ASICs. Recent asymmetric autoencoders deliver high quality under extreme power and bandwidth constraints, but add prohibitive decoding cost and use bespoke formats that ignore decades of infrastructure built around standards like JPEG. To address these limitations, we introduce a compression framework for cloud robotics based on a Sensor Embedded Autoencoder paired with a One-Time Transcode for Efficient Reconstruction (SEAOTTER). Because the sensor, cloud, and consumer stages face very different power and bandwidth budgets, SEAOTTER combines the compactness of a learned latent with the broad usability of a standard JPEG file. Since naive transcoding degrades performance, we propose a learnable JPEG color and quantization transform that enables increased accuracy for global, dense, and vision-language-based perception. Using SEAOTTER, we train both general-purpose and task-aware transcoding pipelines for a pre-trained, frozen encoder. At a compression ratio of 200:1 and compared to AVIF, we observe 7 times faster encoding, 3.5 times faster decoding, and +8% ImageNet top-1 accuracy, while retaining compatibility with JPEG infrastructure. Our code is available at https://github.com/UT-SysML/seaotter .

This letter proposes a novel distributed pose estimator for multi-robot systems evolving on $\mathrm{SE}(3)$. The robots' positions are estimated in $\mathbb{R}^3$ using angles computed from body-frame bearings, without requiring orientation knowledge. The robots' orientations are then recovered in $\mathrm{SO}(3)$ from the estimated positions, together with bearing and bearing-rate measurements. The estimator accommodates directed sensing topologies and requires only infinitesimal angle rigidity (IAR), thereby relaxing the requirement, common in bearing-based approaches, that every robot acquire at least two bearings. Unlike existing angle-based schemes, the proposed method also estimates the robots' orientations. We prove local uniform exponential stability of the observer, assuming that a subset of robots executes persistently exciting motions. These theoretical results are corroborated through numerical simulations.

Motivated by the vision of making sixth-generation (6G) networks sustainable, we study the sparse antenna/array activation problems in uplink cell-free massive multiple-input multiple-output (CF mMIMO) networks. We first develop an antenna-level optimal bilinear equalizer (OBE) weighting framework, in which each access point-user equipment (AP-UE) pair is assigned a matrix-valued long-term weight to shape the contribution of individual antenna elements, thereby generalizing the conventional large-scale fading decoding (LSFD) strategy from scalar coefficients to antenna-element-aware weighting. Building on this structure, we formulate sparse antenna activation as structured sparsity-inducing mean square error (MSE) minimization problems, and design four activation schemes at two granularities: antenna-level and array-level, each with UE-specific and network-wide (all-UEs) variants. The resulting convex problems are solved efficiently via the proximal method with closed-form group-wise updates, while the network-wide schemes are modeled through hierarchical sparsity and handled by a tree-structured proximal operator. Numerical results under correlated Rician channels and a detailed power consumption model demonstrate that the OBE weighting scheme consistently improves spectral efficiency over the LSFD, with gains increasing with the number of antennas. Meanwhile, the studied sparse activation schemes can achieve substantial energy efficiency improvement and power reduction with controllable spectral efficiency loss.

The growing penetration of distributed energy resources (DERs) intensifies congestion in distribution networks by introducing bidirectional power flows and increasing competition for limited network capacity, underscoring the need for effective and efficient congestion management, including flexible grid-access schemes. This paper proposes a bilevel optimization model for the dynamic allocation of connection capacity to DERs under non-firm connection agreements, aligning the objectives of distribution system operator (DSO) and DER owners. The upper-level problem, representing the DSO, determines the allocated connection capacity for all DERs, defined as maximum time-varying power limits, subject to distribution system constraints and the last-in-first-out (LIFO) allocation rule. The lower-level problem, representing DER owners, maximizes the profit of each DER within the allocated power limits. The proposed model is tested on a modified CIGRE medium-voltage (MV) network, demonstrating a balanced trade-off between grid utilization and economic efficiency. Furthermore, the model enhances DER integration, enforces transparent allocation rules, reduces variability in allocation patterns, and achieves up to an 80% reduction in total curtailment costs compared with benchmark methods.

Supervisory control of discrete-event systems provides formal guarantees of correctness with respect to a plant model and specification. However, these guarantees heavily rely on the plant model, which could deviate from nominal behavior due to modeling errors or faults. Recent notions of discrete robustness model deviations as a set of additional transitions that are added to the plant. The discrete robustness is defined as all sets of extra transitions for which the supervised system still guarantees a desired specification. However, this notion suffers from scalability due to the large solution space and conservatism since most deviations are infeasible in practice. This paper proposes to address these two issues using a neurosymbolic computing framework for discrete robustness analysis of safety properties. First, a neural reasoning layer based on Large Language Models infers a set of feasible deviation transitions from system models, specifications, and domain knowledge. Next, a symbolic layer computes the discrete robustness guarantees over the inferred deviation set. We evaluate our framework on three case studies, demonstrating that our method identifies a smaller set of feasible deviations while preserving robustness guarantees comparable to those of full transition-based analysis.

Computing the Lipschitz constant of the solution map of a multi-parametric quadratic program is important for the analysis of optimization-based control. This problem is governed by three factors: the parameter dimension, the number of decision variables, and the number of constraints. While empirical evidence has long suggested exponential complexity, a rigorous complexity-theoretic proof has been lacking. In this paper, we fill this gap by proving that this problem is not only NP-hard but also APX-hard. Furthermore, we reveal that: (a) the problem becomes polynomial-time solvable when the number of constraints or decision variables is fixed; and (b) both NP-hardness and APX-hardness persist even in the scalar parameter case. These results confirm that the complexity stems from the number of constraints and variables, rather than the parameter dimension. Numerical experiments further validate these theoretical findings.

Acoustic feedback limits the maximum gain in hearing aids. In addition to several approaches based on adaptive filtering, recently a deep-neural-network-based feedback cancellation (DFC) approach has been proposed, which is trained via an open-loop framework. Since open-loop-trained DFC (DFC-OL) can become unstable during inference at high gains, in this paper we propose an in-the-loop-trained DFC (DFC-IL) that integrates the DFC directly into the optimisation loop. This allows the model to be exposed to unstable conditions during training. A two-stage training strategy involving pre-training on stable systems and fine-tuning on a wider gain range enables DFC-IL to learn robust howling reduction. Experimental results on measured feedback paths demonstrate that in scenarios with small gains, the proposed DFC-IL performs similarly to DFC-OL, and both exceed the performance of adaptive filters. In scenarios with high amplification gains, DFC-IL clearly outperforms DFC-OL by maintaining system stability.

Pinching antenna systems (PASS) have recently emerged as a promising architecture for flexible indoor wireless communications. However, most existing pinching antenna (PA) array designs for multi-user PASS either offer limited beam adaptation accuracy or require prohibitively high deployment cost. In this paper, we investigate a more practical constrained pinching antenna array (C-PAA)-assisted downlink PASS, where multiple PAs are grouped into a movable array and can be finely adjusted within the array at the wavelength scale. To improve the system spectral efficiency, a sum-rate maximization problem is formulated by jointly considering the array-center position and the fine-grained antenna distribution within the C-PAA. First, the structural properties of the C-PAA are characterized, and an explicit upper bound on the array aperture is derived. Then, tractable approximations for the effective channel gain and the achievable user rate are developed. Furthermore, the optimization problem of the multi-user sum-rate is analyzed, where the system sum-rate function is shown to exhibit a favorable unimodal behavior under practically relevant conditions, which enables an efficient one-dimensional search for the optimal C-PAA position. To further reduce the computational complexity, a closed-form approximate solution for the near-optimal array-center position is derived. Numerical results verify the accuracy of the developed analysis and demonstrate that the proposed C-PAA scheme closely approaches the ideal upper bound and significantly outperforms conventional fixed-spacing and existing PA array benchmarks.

We present LiveBand, a real-time system that generates high-fidelity music accompaniments to live audio input, respecting strict causal constraints. Our method trains a causal transformer generator in the continuous latent space of a pre-trained causal audio autoencoder, using adversarial sequence-level supervision from a discriminator. At each timestep, the generator receives only the causally available mix context and Gaussian noise, and predicts accompaniment latents without access to future mix frames or ground-truth target latents. Training is performed in a single parallel forward pass under causal masking, while streaming inference proceeds autoregressively with a rolling attention state. The model's training and inference computations are matched by design, eliminating teacher forcing and the associated exposure bias. On a multi-instrument music accompaniment benchmark, LiveBand improves over prior work on objective measures of audio quality, beat alignment, and mix adherence, while enabling real-time streaming generation without lookahead into the future on consumer hardware.

Graph Neural Networks (GNNs) have emerged as a powerful tool for wireless resource allocation that leverages the underlying graph structure of communication networks. Their transferability property enables models trained on small-scale graphs to generalize to large-scale deployments with little performance deterioration, a desirable property for currently growing networks. Wireless networks are sparse regimes, where a single node is connected to a small number of other users. This work establishes theoretical results for transferability of GNNs over graphs derived from sparse Random Geometric Graphs (RGGs). In particular, we focus on conflict graphs of RGGs used to model interference among links. Our approach considers the closeness between RGGs and Deterministic Grid Graphs (DGG) to establish bounds in the performance loss when a model is transferred across scales. We validate our theoretical findings through the problem of link scheduling, demonstrating that our learned policies consistently outperform existing benchmarks at scale. Finally, we examine the impact of our theoretical assumptions on empirical performance.

Reliable and affordable electricity supply remains a challenge for remote and regional communities, motivating the deployment of renewable-based microgrids supported by flexible storage and advanced planning methods. This paper proposes an integrated techno-economic framework for optimal microgrid design and robustness assessment, and applies it to a 1000-household residential community in Rockhampton, Queensland (Australia). The framework links time-series simulation, dispatch-based operation, and lifecycle costing to evaluate hybrid configurations comprising photovoltaic and wind generation, battery storage, diesel backup, grid exchange, and an optional hydrogen subsystem (electrolyzer–hydrogen storage–fuel cell). Key indicators include net present cost (NPC), cost of energy (COE), renewable penetration, energy purchased/sold, and emissions-related outcomes. To avoid conclusions that depend on a single set of assumptions, the study performs systematic sensitivity analysis across financial, technical and policy drivers: discount rate, technology capital costs, fuel price, load uncertainty, renewable resource variability, carbon pricing/emissions cost, and grid outage duration, supplemented by a no-hydrogen attribution case. The results demonstrate that several sensitivity dimensions induce nonlinear shifts in the optimal design, including breakpoints where capital-intensive renewable–storage expansion becomes economically preferable. The proposed framework enables transparent comparison of hydrogen-enabled and battery-centric solutions and provides planning guidance for resilient, low-emission community microgrids under Australian operating conditions.

Audio-Visual Event Recognition (AVER) is essential for intelligent urban monitoring systems, where robust multimodal understanding of complex environments is required. This paper proposes a stable hybrid cross-attention fusion framework for audio-visual event recognition in smart urban environments. The proposed architecture combines pretrained Video Masked Autoencoder (VideoMAE) and Audio Spectrogram Transformer (AST) representations with FiLM-based audio conditioning, bidirectional cross-attention fusion, multimodal Transformer encoding, and modality-temporal attention. To improve computational efficiency and training stability, frozen pretrained backbones and cached feature extraction are employed. Extensive experiments on the AVE dataset show that the proposed framework achieves the highest average performance among the evaluated unimodal and multimodal baselines across multiple evaluation metrics, obtaining a best validation accuracy of 91.74% and a test accuracy of 83.85 plus/minus 1.40% over five independent runs. The results indicate that the proposed hybrid fusion strategy effectively captures complementary audio-visual information and provides robust multimodal representation learning for challenging realworld urban monitoring scenarios.

We demonstrated hybrid free-space optics (FSO) and D-band (110-170GHz) millimetre wave transmitter enabled by a single phase-locked laser pair, simultaneously enabling ultra-low RF phase noise and optical linewidth for communications. Based on this, we further study combined capacity with beam angle misalignment using >100Gb/s signalling.

While the hybrid energy storage system (HESS) can theoretically mitigate battery degradation in electric vehicles, its practical implementation remains highly limited. To delineate the specific scenarios and application boundaries where supercapacitors remain feasible, this study proposes a multi-dimensional techno-economic feasibility evaluation framework. First, a cross-vehicle sizing method based on dynamic programming is established to quantify physical mass-volume packaging constraints and identify feasible supercapacitor candidates across different vehicle types. Building upon the optimal sizing parameters derived from the battery aging Pareto front, an expert-guided deep reinforcement learning energy management strategy is integrated to yield near-optimal online performance, ensuring a fair life-cycle economic assessment. Finally, a comprehensive feasibility matrix is constructed to systematically evaluate mass, volume, battery lifespan, additional supercapacitor costs, operation cost, future energy storage prices, and the influence of emerging solid-state batteries. Results reveal that city buses remain the most promising vehicle type for HESS due to minimal additional costs and sufficient packaging space. Current mass-volume penalties and limited economic benefits hinder HESS application in passenger vehicles and heavy-duty trucks, respectively. This situation may only improve if supercapacitor prices drop significantly in the future. Quantified analysis shows that higher load-frequency characteristics generally favor HESS benefits within the same vehicle platform, while the overall techno-economic feasibility across vehicles is jointly influenced by load characteristics and vehicle configuration. Furthermore, looking toward the 2030+ solid-state battery era, we highlight that integrating increasingly affordable supercapacitors can provide substantial asset protection leverage.