Papers

In many sensing (viz., radio astronomy) and radar applications, the received signal of interest (SOI) exhibits a significantly wider bandwidth or weaker power than the interference signal, rendering it indistinguishable from the background noise. Such scenarios arise frequently in applications such as passive radar, cognitive radio, low-probability-of-intercept (LPI) radar, and planetary radar for radio astronomy, where canceling the radio frequency interference (RFI) is critical for uncovering the SOI. In this work, we examine the Demodulation-Remodulation (Demod-Remod) based interference cancellation framework for the RFI. This approach demodulates the unknown interference, creates a noise-free interference replica, and coherently subtracts it from the received signal. To evaluate the performance limits, we employ the performance metric termed interference rejection ratio (IRR), which quantifies the interference canceled. We derive the analytical expressions of IRR as a function of the optimal estimation variances of the signal parameters. Simulation results confirm the accuracy of the analytical expression for both single-carrier and multi-carrier interference signals and demonstrate that the method can substantially suppress the interference at a sufficient interference-to-noise ratio (INR), enabling enhanced detection and extraction of the SOI. We further extend the analysis to the scenario where the SOI is above the noise floor, and confirm the validity of the theoretical IRR expression in this scenario. Lastly, we compare the Demod-Remod technique to other time-domain cancellation methods. The result of the comparison identifies the conditions under which each method is preferred, offering practical guidelines for interference mitigation under different scenarios.

Reconfigurable Intelligent Surfaces (RIS) have recently gained attention as a means to dynamically shape the wireless propagation environment through programmable reflection control. Among the numerous applications, an important emerging use case is employing RIS as an auxiliary mechanism for spatial interference nulling, particularly in large ground-based reflector antennas where sidelobe interference can significantly degrade the system performance. With the growing density of satellites and terrestrial emitters, algorithms with faster convergence speed and better performance are needed. This work investigates RIS-equipped reflector antennas as a representative example of RIS-assisted spatial nulling and develop algorithms for sidelobe cancellation at specific directions and frequencies under various constraints. For the continuous-phase case, we adapt the gradient projection (GP) and alternating projection (AP) algorithms for scalability and propose a closed-form near-optimal solution that achieves satisfactory nulling performance with significantly reduced complexity. For the discrete-phase case, we reformulate the problem using a penalty method and solve it via majorization-minimization, outperforming the heuristic methods from our earlier work. Further, we analyze the electric field characteristics across multiple interference directions and frequencies to quantify the nulling capability of the RIS-aided reflectors, and identify a simple criterion for the existence of unimodular weights enabling perfect nulls. Simulation results demonstrate the effectiveness of the proposed methods and confirm the theoretical nulling limits.

Several optimism-based stochastic bandit algorithms – including UCB, UCB-V, linear UCB, and finite-arm GP-UCB – achieve logarithmic regret using proofs that, despite superficial differences, follow essentially the same structure. This note isolates the minimal ingredients behind these analyses: a single high-probability concentration condition on the estimators, after which logarithmic regret follows from two short deterministic lemmas describing radius collapse and optimism-forced deviations. The framework yields unified, near-minimal proofs for these classical algorithms and extends naturally to many contemporary bandit variants.

This informal contribution presents an ongoing line of research that is pursuing a new approach to the construction of sound proofs for the formal verification and control of complex stochastic models of dynamical systems, of reactive programs and, more generally, of models of Cyber-Physical Systems. Neural proofs are made up of two key components: 1) proof rules encode requirements entailing the verification of general temporal specifications over the models of interest; and 2) certificates that discharge such rules, namely they are constructed from said proof rules with an inductive (that is, cyclic, repetitive) approach; this inductive approach involves: 2a) accessing samples from the model's dynamics and accordingly training neural networks, whilst 2b) generalising such networks via SAT-modulo-theory (SMT) queries that leverage the full knowledge of the models. In the context of sequential decision making problems over complex stochastic models, it is possible to additionally generate provably-correct policies/strategies/controllers, namely state-feedback functions that, in conjunction with neural certificates, formally attain the given specifications for the models of interest.

Recently, the Large Language Model-based Phoneme-to-Grapheme (LLM-P2G) method has shown excellent performance in speech recognition tasks and has become a feasible direction to replace the traditional WFST decoding method. This framework takes into account both recognition accuracy and system scalability through two-stage modeling of phoneme prediction and text generation. However, the existing LLM-P2G adopts the Top-K Marginalized (TKM) training strategy, and its candidate phoneme sequences rely on beam search generation, which has problems such as insufficient path diversity, low training efficiency, and high resource overhead. To this end, this paper proposes a sampling marginalized training strategy (Sampling-K Marginalized, SKM), which replaces beam search with random sampling to generate candidate paths, improving marginalized modeling and training efficiency. Experiments were conducted on Polish and German datasets, and the results showed that SKM further improved the model learning convergence speed and recognition performance while maintaining the complexity of the model. Comparative experiments with a speech recognition method that uses a projector combined with a large language model (SpeechLLM) also show that the SKM-driven LLM-P2G has more advantages in recognition accuracy and structural simplicity. The study verified the practical value and application potential of this method in cross-language speech recognition systems.

Diffusion models are highly expressive image priors for Bayesian inverse problems. However, most diffusion models cannot operate on large-scale, high-dimensional data due to high training and inference costs. In this work, we introduce a Plug-and-play algorithm for 3D stochastic inference with latent diffusion prior (PSI3D) to address massive ($1024\times 1024\times 128$) volumes. Specifically, we formulate a Markov chain Monte Carlo approach to reconstruct each two-dimensional (2D) slice by sampling from a 2D latent diffusion model. To enhance inter-slice consistency, we also incorporate total variation (TV) regularization stochastically along the concatenation axis. We evaluate our performance on optical coherence tomography (OCT) super-resolution. Our method significantly improves reconstruction quality for large-scale scientific imaging compared to traditional and learning-based baselines, while providing robust and credible reconstructions.

Conventional robust H2/H-infinity control minimizes the worst-case performance, often leading to a conservative design driven by very rare parametric configurations. To reduce this conservatism while taking advantage of the stochastic properties of Monte Carlo sampling and its compatibility with parallel computing, we introduce an alternative paradigm that optimizes the controller with respect to a stochastic criterion, namely the conditional value at risk. We present the problem formulation and discuss several open challenges toward a general synthesis framework. The potential of this approach is illustrated on a mechanical system, where it significantly improves overall performance by tolerating some degradation in very rare worst-case scenarios.

Sentiment analysis using Electroencephalography (EEG) sensor signals provides a deeper behavioral understanding of a person's emotional state, offering insights into real-time mood fluctuations. This approach takes advantage of brain electrical activity, making it a promising tool for various applications, including mental health monitoring, affective computing, and personalised user experiences. An encoder-based model for EEG-to-sentiment analysis, utilizing the ZUCO 2.0 dataset and incorporating a Feature Pyramid Network (FPN), is proposed to enhance this process. FPNs are adapted here for EEG sensor data, enabling multiscale feature extraction to capture local and global sentiment-related patterns. The raw EEG sensor data from the ZUCO 2.0 dataset is pre-processed and passed through the FPN, which extracts hierarchical features. In addition, extracted features are passed to a Gated Recurrent Unit (GRU) to model temporal dependencies, thereby enhancing the accuracy of sentiment classification. The ZUCO 2.0 dataset is utilized for its clear and detailed representation in 128 channels, offering rich spatial and temporal resolution. The experimental metric results show that the proposed architecture achieves a 6.88% performance gain compared to the existing methods. Furthermore, the proposed framework demonstrated its efficacy on the validation datasets DEAP and SEED.

This paper presents an operator-theoretic reconstruction of an equivalent continuous-time LTI model from an exact sampled-data (Poincaré-map) baseline of a piecewise-linear switching system. The rebuilding is explicitly expressed via matrix logarithms. By expanding the logarithm of a product of matrix exponentials using the Baker–Campbell–Hausdorff (BCH) formula, we show that the classical state-space averaging (SSA) model can be interpreted as the leading-order truncation of this exact reconstruction when the switching period is small and the ripple is small. The same view explains why SSA critically relies on low-frequency and small-ripple assumptions, and why the method becomes fragile for converters with more than two subintervals per cycle. Finally, we provide a complexity-reduced, SSA-flavoured implementation strategy for obtaining the required spectral quantities and a real-valued logarithm without explicitly calling eigen-decomposition or complex matrix logarithms, by exploiting $2\times 2$ invariants and a minimal real-lift construction.

Operating power amplifiers (PAs) at lower input back-off (IBO) levels is an effective way to improve PA efficiency, but often introduces severe nonlinear distortion that degrades transmission performance. Amplitude-phase-time block modulation (APTBM) has recently emerged as an effective solution to this problem. The intrinsic amplitude and phase constraints of each APTBM block can be leveraged to mitigate PA-induced nonlinear distortion via constraint-guided signal reconstruction. However, existing reconstruction methods apply these constraints only heuristically and statistically, limiting the achievable IBO reduction and PA efficiency improvement. This paper addresses this limitation by decomposing the nonlinear distortion into dominant and residual components, and accordingly develops a novel two-stage signal reconstruction algorithm consisting of coarse and fine reconstruction stages. The coarse reconstruction stage eliminates the dominant distortion by jointly exploiting the APTBM block structure and PA nonlinear characteristics. Subsequently, the fine reconstruction stage minimizes the residual distortion by casting it as a nonconvex optimization problem subject to explicit APTBM constraints, for which a closed-form solution is derived. The proposed algorithm is validated through comprehensive numerical simulations and testbed experiments. Results show that, without compromising transmission quality, the proposed algorithm enables an additional IBO reduction of approximately 5 dB in simulations and 2 dB in experiments over baseline methods, yielding relative PA efficiency improvements of 77.8% and 30.9%, respectively.

This paper presents a trustworthy reinforcement learning approach for the control of industrial compressed air systems. We develop a framework that enables safe and energy-efficient operation under realistic boundary conditions and introduce a multi-level explainability pipeline combining input perturbation tests, gradient-based sensitivity analysis, and SHAP (SHapley Additive exPlanations) feature attribution. An empirical evaluation across multiple compressor configurations shows that the learned policy is physically plausible, anticipates future demand, and consistently respects system boundaries. Compared to the installed industrial controller, the proposed approach reduces unnecessary overpressure and achieves energy savings of approximately 4\,% without relying on explicit physics models. The results further indicate that system pressure and forecast information dominate policy decisions, while compressor-level inputs play a secondary role. Overall, the combination of efficiency gains, predictive behavior, and transparent validation supports the trustworthy deployment of reinforcement learning in industrial energy systems.

The mitigation of unbalanced grid voltages or currents by voltage source converters results in power ripple on the dc link, and is a key converter design parameter due to hardware or stability considerations. Despite the importance of this issue for system design and operation, the use of Optimal Power Flow (OPF)-based methods capturing the interaction between dc link ripple and converter unbalanced operation has been largely unexplored. In this work, the magnitude of the power ripple is derived for generic multi-terminal converters, then introduced as a bilinear OPF constraint for two-level converter topologies. OPF case studies demonstrate the necessity to model both neutral current and dc link ripple, with tradeoffs between capacitor sizing and leg sizing highlighted for phase current unbalance mitigation applications. Time domain simulations of a grid-connected four-wire voltage source converter verify the accuracy and validity of the algebraic formulation. It is concluded that awareness of dc link ripple impacts and constraints will be of growing importance for distribution system operators.

Developing a good speaker embedding has received tremendous interest in the speech community, with representations such as i-vector and d-vector demonstrating remarkable performance across various tasks. Despite their widespread adoption, a fundamental question remains largely unexplored: what properties are actually encoded in these embeddings? To address this gap, we conduct a comprehensive analysis of three prominent speaker embedding methods: i-vector, d-vector, and RNN/LSTM-based sequence-vector (s-vector). Through carefully designed classification tasks, we systematically investigate their encoding capabilities across multiple dimensions, including speaker identity, gender, speaking rate, text content, word order, and channel information. Our analysis reveals distinct strengths and limitations of each embedding type: i-vector excels at speaker discrimination but encodes limited sequential information; s-vector captures text content and word order effectively but struggles with speaker identity; d-vector shows balanced performance but loses sequential information through averaging. Based on these insights, we propose a novel multi-task learning framework that integrates i-vector and s-vector, resulting in a new speaker embedding (i-s-vector) that combines their complementary advantages. Experimental results on RSR2015 demonstrate that the proposed i-s-vector achieves more than 50% EER reduction compared to the i-vector baseline on content mismatch trials, validating the effectiveness of our approach.

Children's speech recognition remains challenging due to substantial acoustic and linguistic variability, limited labeled data, and significant differences from adult speech. Speech foundation models can address these challenges through Speech In-Context Learning (SICL), allowing adaptation to new domains without fine-tuning. However, the effectiveness of SICL depends on how in-context examples are selected. We extend an existing retrieval-based method, Text-Embedding KNN for SICL (TICL), introducing an acoustic reranking step to create TICL+. This extension prioritizes examples that are both semantically and acoustically aligned with the test input. Experiments on four children's speech corpora show that TICL+ achieves up to a 53.3% relative word error rate reduction over zero-shot performance and 37.6% over baseline TICL, highlighting the value of combining semantic and acoustic information for robust, scalable ASR in children's speech.

This letter investigates the coupled control problem in UAV networks utilizing high-frequency hybrid beamsteering. While phased arrays enable rapid electronic scanning, their finite Field of View (FoV) imposes a fundamental constraint that necessitates active mechanical steering of the airframe to maintain connectivity. We propose a decentralized Model Predictive Control (MPC) framework that jointly optimizes trajectory and heading to maximize network sum-capacity subject to safety constraints. Addressing the numerical instability caused by fast-fading channel nulls, we introduce a regularized surrogate cost function based on discrete spatial smoothing. We analytically prove that this approximation bounds the cost curvature, restoring the Lipschitz continuity of the gradient. Crucially, we derive a sufficient condition linking this Lipschitz constant to the controller gain, guaranteeing the contraction and linear convergence of the distributed best-response dynamics. Simulation results demonstrate that the proposed algorithm effectively navigates the trade-off between electronic beam tracking and kinematic safety, significantly systematically outperforming velocity-aligned baselines.

Achieving robust generalization in speech deepfake detection (SDD) remains a primary challenge, as models often fail to detect unseen forgery methods. While research has focused on model-centric and algorithm-centric solutions, the impact of data composition is often underexplored. This paper proposes a data-centric approach, analyzing the SDD data landscape from two practical perspectives: constructing a single dataset and aggregating multiple datasets. To address the first perspective, we conduct a large-scale empirical study to characterize the data scaling laws for SDD, quantifying the impact of source and generator diversity. To address the second, we propose the Diversity-Optimized Sampling Strategy (DOSS), a principled framework for mixing heterogeneous data with two implementations: DOSS-Select (pruning) and DOSS-Weight (re-weighting). Our experiments show that DOSS-Select outperforms the naive aggregation baseline while using only 3% of the total available data. Furthermore, our final model, trained on a 12k-hour curated data pool using the optimal DOSS-Weight strategy, achieves state-of-the-art performance, outperforming large-scale baselines with greater data and model efficiency on both public benchmarks and a new challenge set of various commercial APIs.

Current data-driven Wi-Fi-based indoor localization systems face three critical challenges: protecting user privacy, achieving accurate predictions in dynamic multipath environments, and generalizing across different deployments. Traditional Wi-Fi localization systems often compromise user privacy, particularly when facing compromised access points (APs) or man-in-the-middle attacks. As IoT devices proliferate in indoor environments, developing solutions that deliver accurate localization while robustly protecting privacy has become imperative. We introduce FedWiLoc, a privacy-preserving indoor localization system that addresses these challenges through three key innovations. First, FedWiLoc employs a split architecture where APs process Channel State Information (CSI) locally and transmit only privacy-preserving embedding vectors to user devices, preventing raw CSI exposure. Second, during training, FedWiLoc uses federated learning to collaboratively train the model across APs without centralizing sensitive user data. Third, we introduce a geometric loss function that jointly optimizes angle-of-arrival predictions and location estimates, enforcing geometric consistency to improve accuracy in challenging multipath conditions. Extensive evaluation across six diverse indoor environments spanning over 2,000 sq. ft. demonstrates that FedWiLoc outperforms state-of-the-art methods by up to 61.9% in median localization error while maintaining strong privacy guarantees throughout both training and inference.

In recent years, the demand of image compression models for machine vision has increased dramatically. However, the training frameworks of image compression still focus on the vision of human, preserving unnecessary details and limiting bitrate efficiency for machine vision tasks. In this paper, we propose Semantic-based Low-bitrate Image compression for Machines, termed SLIM. This proposed framework consists of two key components: RoI-focused latent compression and semantic latent enhancement with dual cues. First, the latent compressor is trained to focus on Region-of-Interest (RoI) areas that are semantically and functionally important for machine vision tasks in the latent space, while requiring no explicit masks during inference. Second, to compensate for severe information loss at extremely low bitrates, a diffusion-based semantic decoder enhances the compressed latent with a pretrained diffusion model guided by two complementary cues provided by an RoI-focused latent and text-based semantic guidance. This design preserves RoI details that are critical for machine perception while maintaining semantically coherent and artifact-suppressed non-RoI regions under extremely low bitrates. Experimental results show that SLIM consistently achieves strong performance across diverse machine vision tasks under the same bits per pixel condition, compared to conventional image compression models for machines. Our code is available at https://github.com/mcml-official/SLIM

Perivascular spaces (PVS), when abnormally enlarged and visible in magnetic resonance imaging (MRI) structural sequences, are important imaging markers of cerebral small vessel disease and potential indicators of neurodegenerative conditions. Despite their clinical significance, automatic enlarged PVS (EPVS) segmentation remains challenging due to their small size, variable morphology, similarity with other pathological features, and limited annotated datasets. This paper presents the EPVS Challenge organized at MICCAI 2024, which aims to advance the development of automated algorithms for EPVS segmentation across multi-site data. We provided a diverse dataset comprising 100 training, 50 validation, and 50 testing scans collected from multiple international sites (UK, Singapore, and China) with varying MRI protocols and demographics. All annotations followed the STRIVE protocol to ensure standardized ground truth and covered the full brain parenchyma. Seven teams completed the full challenge, implementing various deep learning approaches primarily based on U-Net architectures with innovations in multi-modal processing, ensemble strategies, and transformer-based components. Performance was evaluated using dice similarity coefficient, absolute volume difference, recall, and precision metrics. The winning method employed MedNeXt architecture with a dual 2D/3D strategy for handling varying slice thicknesses. The top solutions showed relatively good performance on test data from seen datasets, but significant degradation of performance was observed on the previously unseen Shanghai cohort, highlighting cross-site generalization challenges due to domain shift. This challenge establishes an important benchmark for EPVS segmentation methods and underscores the need for the continued development of robust algorithms that can generalize in diverse clinical settings.

Lithium-ion batteries are the major type of battery used in a variety of everyday applications, including electric vehicles (EVs), mobile devices, and energy storage systems. Predicting the Remaining Useful Life (RUL) of Li-ion batteries is crucial for ensuring their reliability, safety, and cost-effectiveness in battery-powered systems. The materials used for the battery cathodes and their designs play a significant role in determining the degradation rates and RUL, as they lead to distinct electrochemical reactions. Unfortunately, RUL prediction models often overlook the cathode materials and designs to simplify the model-building process, ignoring the effects of these electrochemical reactions. Other reasons are that specifications related to battery materials may not always be readily available, and a battery might consist of a mix of different materials. As a result, the predictive models that are developed often lack generalizability. To tackle these challenges, this paper proposes a novel material-based Mixture-of-Experts (MoE) approach for predicting the RUL of batteries, specifically addressing the complexities associated with heterogeneous battery chemistries. The MoE is integrated into a probabilistic framework, called Multiple Non-crossing Quantile Mixture-of-Experts for Probabilistic Prediction (RUL-QMoE), which accommodates battery operational conditions and enables uncertainty quantification. The RUL-QMoE model integrates specialized expert networks for five battery types: LFP, NCA, NMC, LCO, and NMC-LCO, within a gating mechanism that dynamically assigns relevance based on the battery's input features. Furthermore, by leveraging non-crossing quantile regression, the proposed RUL-QMoE produces coherent and interpretable predictive distributions of the battery's RUL, enabling robust uncertainty quantification in the battery's RUL prediction.

Timely brain tumor diagnosis remains challenging in low-resource clinical environments where expert neuroradiology interpretation, high-end MRI hardware, and invasive biopsy procedures may be limited. Although deep learning has achieved strong performance in brain tumor analysis, real-world adoption is constrained by computational demands, dataset shift across scanners, and limited interpretability. This paper presents a prototype virtual biopsy pipeline for four-class classification of 2D brain MRI images using a lightweight convolutional neural network (CNN) and complementary radiomics-style handcrafted features. A MobileNetV2-based CNN is trained for classification, while an interpretable radiomics branch extracts eight features capturing lesion shape, intensity statistics, and gray-level co-occurrence matrix (GLCM) texture descriptors. A late fusion strategy concatenates CNN embeddings with radiomics features and trains a RandomForest classifier on the fused representation. Explainability is provided via Grad-CAM visualizations and radiomics feature importance analysis. Experiments on a public Kaggle brain tumor MRI dataset show improved validation performance for fusion relative to single-branch baselines, while robustness tests under reduced resolution and additive noise highlight sensitivity relevant to low-resource imaging conditions. The system is framed as decision support and not a substitute for clinical diagnosis or histopathology.

The increasing use of electric vehicles (EVs) requires efficient route planning solutions that take into account the limited range of EVs and the associated charging times, as well as the different types of charging stations. In this work, we model and solve an electric vehicle routing problem (EVRP) designed for a cross-platform navigation system for individual transport. The aim is to provide users with an efficient route for their daily appointments and to reduce possible inconveniences caused by charging their EV. Based on these assumptions, we propose a multi-day model in the form of a mixed integer programming (MIP) problem that takes into account the vehicle's battery capacity and the time windows of user's appointments. The model is solved using various established metaheuristics, including tabu search (TS), adaptive large neighborhood search (ALNS), and ant colony optimization (ACO). Furthermore, the performance of the individual approaches is analyzed using generated ensembles to estimate their behavior in reality and is compared with the exact results of the Google OR-Tools solver.

This paper addresses the challenge of modeling and control in hierarchical, multi-agent systems, known as holonic systems, where local agent decisions are coupled with global systemic outcomes. We introduce the Bayesian Holonic Equilibrium (BHE), a concept that ensures consistency between agent-level rationality and system-wide emergent behavior. We establish the theoretical soundness of the BHE by showing its existence and, under stronger regularity conditions, its uniqueness. We propose a two-time scale learning algorithm to compute such an equilibrium. This algorithm mirrors the system's structure, with a fast timescale for intra-holon strategy coordination and a slow timescale for inter-holon belief adaptation about external risks. The convergence of the algorithm to the theoretical equilibrium is validated through a numerical experiment on a continuous public good game. This work provides a complete theoretical and algorithmic framework for the principled design and analysis of strategic risk in complex, coupled control systems.

We study self-triggered two-player stochastic games on Piecewise Deterministic Markov Processes (PDMPs) where each agent decides when to observe and which open-loop action to hold. Augmenting the state with clocks and committed controls yields flow regions (both hold) and trigger surfaces (at least one updates). The framework covers both blind simultaneous (Nash) timing and observable sequential (Stackelberg) commitments; the former leads to coupled, intractable QVIs, while the latter admits a nested Hamilton-Jacobi-Bellman quasi-variational inequality and a tractable dynamic-programming decomposition. We outline a computational scheme based on implicit differentiation of the follower's fixed point. A pursuit-evasion example illustrates the strategic timing interaction.

General audio source separation is a key capability for multimodal AI systems that can perceive and reason about sound. Despite substantial progress in recent years, existing separation models are either domain-specific, designed for fixed categories such as speech or music, or limited in controllability, supporting only a single prompting modality such as text. In this work, we present SAM Audio, a foundation model for general audio separation that unifies text, visual, and temporal span prompting within a single framework. Built on a diffusion transformer architecture, SAM Audio is trained with flow matching on large-scale audio data spanning speech, music, and general sounds, and can flexibly separate target sources described by language, visual masks, or temporal spans. The model achieves state-of-the-art performance across a diverse suite of benchmarks, including general sound, speech, music, and musical instrument separation in both in-the-wild and professionally produced audios, substantially outperforming prior general-purpose and specialized systems. Furthermore, we introduce a new real-world separation benchmark with human-labeled multimodal prompts and a reference-free evaluation model that correlates strongly with human judgment.

This paper develops a hierarchical games-in-games control architecture for hybrid stochastic systems governed by regime-switching jump-diffusions. We model the interplay between continuous state dynamics and discrete mode transitions as a bilevel differential game: an inner layer solves a robust stochastic control problem within each regime, while a strategic outer layer modulates the transition intensities of the underlying Markov chain. A Dynkin-based analysis yields a system of coupled Hamilton-Jacobi-Isaacs (HJI) equations. We prove that for the class of Linear-Quadratic games and Exponential-Affine games, this hierarchy admits tractable semi-closed form solutions via coupled matrix differential equations. We prove that for the class of Linear-Quadratic games and Exponential-Affine games, this hierarchy admits tractable semi-closed form solutions via coupled matrix differential equations. The framework is demonstrated through a case study on adversarial market microstructure, showing how the outer layer's strategic switching pre-emptively adjusts inventory spreads against latent regime risks, which leads to a hyper-alert equilibrium.

Continuous-variable quantum key distribution (CVQKD) over free-space optical links is a promising approach for secure communication, but its performance is limited by turbulence, pointing errors, and angular leakage that can be exploited by an eavesdropper. To mitigate this, we consider an angular rejection filter that defines a safe-zone at the receiver and blocks signals from outside the desired cone. A system and channel model is developed including turbulence, misalignment, and safe-zone effects, and information theoretic metrics are derived to evaluate security. Simulation results show that the safe zone significantly reduces information leakage and that careful tuning of beam waist, angular threshold, and aperture size is essential for maximizing the secret key rate. Larger apertures improve performance but increase receiver size, while longer links require sub 100 urad alignment accuracy. These results highlight safe-zone enforcement and parameter optimization as effective strategies for practical and secure CV-QKD.

We consider outdoor optical access points (OAPs), which, enabled by recent advances in metasurface technology, have attracted growing interest. While OAPs promise high data rates and strong physical-layer security, practical deployments still expose vulnerabilities and misuse patterns that necessitate a dedicated monitoring layer - the focus of this work. We therefore propose a user positioning and monitoring system that infers locations from spatial intensity measurements on a photodetector (PD) array. Specifically, our hybrid approach couples an optics-informed forward model and sparse, model-based inversion with a lightweight data-driven calibration stage, yielding high accuracy at low computational cost. This design preserves the interpretability and stability of model-based reconstruction while leveraging learning to absorb residual nonidealities and device-specific distortions. Under identical hardware and training conditions (both with 5 x 10^5 samples), the hybrid method attains consistently lower mean-squared error than a generic deep-learning baseline while using substantially less training time and compute. Accuracy improves with array resolution and saturates around 60 x 60-80 x 80, indicating a favorable accuracy-complexity trade-off for real-time deployment. The resulting position estimates can be cross-checked with real-time network logs to enable continuous monitoring, anomaly detection (e.g., potential eavesdropping), and access control in outdoor optical access networks.

Pinching-antenna system (PASS) mitigates large-scale path loss by enabling flexible placement of pinching antennas (PAs) along the dielectric waveguide. However, most existing studies assume perfect channel state information (CSI), overlooking the impact of channel uncertainty. This paper addresses this gap by proposing a robust beamforming framework for both lossy and lossless waveguides. For baseband beamforming, the lossy case yields an second-order cone programming-based solution, while the lossless case admits a closed-form solution via maximum ratio transmission. The PAs' positions in both cases are optimized through the Gauss-Seidel-based method. Numerical results validate the effectiveness of the proposed algorithm and demonstrate that PASS exhibits superior robustness against channel uncertainty compared with conventional fixed-antenna systems. Notably, its worst-case achievable rate can even exceed the fixed-antenna baseline under perfect CSI.

Accurate channel impulse response (CIR) modeling in molecular communication (MC) often requires solving coupled reactive diffusion-advection equations, which is computationally expensive for large parameter sweeps or design loops. We develop a deep-learning surrogate for a three-dimensional duct MC channel with reactive diffusion-advection transport and reversible ligand-receptor binding on a finite ring receiver. Using a physics-based partial differential equation (PDE)-ordinary differential equation (ODE) model, we generate a large CIR dataset across broad transport, reaction, and geometric ranges and train a neural network that maps these parameters directly to the CIR. On an independent test set, the surrogate closely matches reference CIRs both qualitatively and quantitatively: the empirical cumulative distribution function (CDF) of the normalized root mean square error (NRMSE) shows that 90% of test channels are predicted with error below 0.15, with only weak dependence on individual parameters. The surrogate therefore offers an accurate and computationally efficient replacement for repeated PDE-based CIR evaluations in MC system analysis and design.