Papers

We present the first multi-subcarrier massive multi-user (MU) multiple-input multiple-output (MIMO) orthogonal frequency-division multiplexing (OFDM) data detector reported in the open literature. By exploiting the channel's beamspace sparsity and frequency-domain (FD) correlation, we achieve up to 3x area and power reduction. Our design supports U=8 user equipments and B=64 basestation antennas, computes soft outputs for QPSK to 256-QAM, and processes 16 subcarriers in parallel. The fabricated 22FDX ASIC has a core cell area of 1.3mm$^2$, consumes 286mW, and delivers a throughput of 7.3Gbps at 0.8V core voltage, achieving best-in-class energy efficiency of 39pJ/b.

Resilience has emerged as a fundamental capability for future wireless systems operating in dynamic and uncertain environments. Although resilience has attracted growing attention across academia, industry, and standardization, its conceptual scope, enabling mechanisms, and realization techniques remain fragmented. This paper presents a systematic framework of resilience in wireless systems from a trustworthiness perspective. We first formalize the concept of resilience and distinguish it from related uncertainty-aware terminologies, including reliability, robustness, adaptability, survivability, and recoverability. We then establish a hierarchical framework that organizes resilience into capability dimensions and enabling mechanisms, and that quantifies resilience through different technical aspects. We further use physical links and unmanned aerial vehicle networks as representative wireless scenarios to demonstrate how resilience can be systematically realized through the joint design of architectures, operations, and algorithms. Finally, we examine the fundamental trade-offs in resilience engineering: Improving resilience generally incurs costs in resource efficiency, nominal performance, information acquisition, implementation complexity, and latency.

This paper proposes a new technique for computer simulation of linear filters. It allows simulating continuous-time linear filters based on the spectral method for analyzing linear control systems. Applying the spectral method implies that the input and output signals are represented by ordered sets of expansion coefficients in a chosen basis, and the filter itself is specified by a two-dimensional nonstationary transfer function. The described technique is tested on Bessel, Papoulis, and Legendre filters of various orders. For each filter, the corresponding two-dimensional nonstationary transfer function, i.e., the matrix of the linear transformation relating expansion coefficients of the input and output signals, is obtained.

Full-waveform inversion (FWI) is a promising strategy for quantitative musculoskeletal ultrasound computed tomography (USCT), but bone-related scattering, attenuation, and signal degradation make it highly sensitive to the accuracy of the initial acoustic-property distributions and prone to cycle skipping. First-arrival traveltimes provide important kinematic information for initial-model construction, yet conventional trace-wise picking is unreliable when arrivals are weak, spatially heterogeneous, or buried in system noise. We propose a learning-assisted reconstruction pipeline that combines segmentation-based first-arrival extraction with hybrid full-waveform inversion (HFWI), which incorporates Rytov-approximation-based traveltime information together with waveform fitting during the early inversion stage. A lightweight 2D U-Net treats the first-arrival trajectory across receiver channels as a first-break segmentation target and exploits its spatial continuity rather than processing each trace independently. To address both limited manual annotations and the simulation-to-real gap, the network is pretrained on task-specific simulations augmented with real system-noise recordings, followed by stage-wise training with progressively increased signal degradation and decoder-only fine-tuning using limited weakly labeled experimental data. The method is evaluated on in vitro phantom, ex vivo bovine-limb, and in vivo human-thigh datasets. Compared with conventional STA/LTA picking, the proposed network yields more spatially coherent first-arrival trajectories, lower mean extraction errors, and processes a full-matrix-capture dataset within seconds. When integrated into HFWI, the extracted arrivals improve initial-model construction and lead to stable subsequent FWI reconstructions, including challenging cases with estimated local first-arrival SNRs below 3 dB.

We compare gradient-based inverse design with structured random search for dielectric-loaded microwave cavities using an in-house JAX-based differentiable FDTD solver. The benchmark fixes material fraction, filtered density representation, initialization, band-energy objective, and measured selection wall time. Each selected design is evaluated in a separate 8000-step FDTD simulation. Across four target bands, three prescribed material fractions, and six seeds, gradient optimization achieves a higher in-band spectral-energy fraction in all 72 paired comparisons. The mean paired improvement is 0.356, with a 95 percent paired bootstrap interval of 0.339-0.374 and an exact two-sided sign-flip p value of 0.03125. Gradient, cavity-mode, CFL, material-fraction, and raw-output integrity checks pass before inference. Under matched material and computational budgets, the results support an advantage for gradient-based optimization in this cavity-design benchmark.

This paper proposes a differentially private federated learning (FL) framework built upon an FL algorithm with semantic feature reconstruction (FedSFR) for training semantic communication modules for image transmission. By allowing clients with unfavorable uplink capacity to transmit low-dimensional semantic feature vectors extracted from locally trained joint source-channel coding (JSCC) encoders, FedSFR enhances communication efficiency and training stability under heterogeneous wireless conditions. To protect client privacy, we incorporate the oneshot Laplace mechanism and theoretically demonstrate that feature-based transmission achieves strictly stronger differential privacy (DP) guarantees than gradient-based transmission under an identical communication budget. In addition, a model selection mechanism is introduced to alleviate performance degradation caused by privacy-preserving perturbations. Experimental results on multiple datasets show that the proposed DP-aided FedSFR outperforms DP-enabled FedAvg in training stability and image reconstruction quality in heterogeneous wireless systems.

This paper investigates range–Doppler (RD) sidelobe suppression in an integrated sensing and communications system with a reconfigurable holographic surface (RHS). We jointly design the digital feed precoders and RHS amplitudes to minimize the integrated RD sidelobe level subject to transmit-power, target-illumination, and communication constraints. For this, we develop an alternating successive convex approximation method updating both variable blocks through quadratic subproblems. Numerical results reveal diminishing returns from additional feeds, while increasing the aperture remains more effective and allows the RHS design to reduce sidelobe level.

The transition to 6G-and-beyond wireless systems with large-scale antenna arrays and high-frequency deployments significantly extends the near-field region, where channels exhibit a strong dependence on user location. While this enables location-based beam focusing as a low-overhead alternative to conventional channel estimation, its performance is highly sensitive to localization errors. In this paper, we study robust near-field beam focusing under imperfect user localization. We explicitly characterize the impact of localization errors on the line-of-sight-dominated channel by deriving a tractable uncertainty model via a first-order Taylor approximation, which captures the coupled effects of distance and angle in near-field propagation. Building on this model, we formulate a max-min signal-to-interference-plus-noise ratio optimization problem that guarantees performance under worst-case channel realizations induced by bounded localization errors. The resulting problem is reformulated into a feasibility problem using semidefinite relaxation. Numerical results demonstrate that the proposed robust design significantly improves the worst-user rate compared to non-robust beam focusing, particularly under high total transmit power levels and large localization error ranges.

Long Short-Term Memory (LSTM) networks continue to be widely deployed for real-time sequential tasks on edge devices, yet their computational demands challenge deployment on resource-constrained hardware. This work introduces COBALT, a bit-serial compressed LSTM accelerator built on a matrix (input)-vector (weight) reformulation of the standard circulant matrix-vector multiplication (MVM) with offset-binary coding. A novel column-swapping scheme operates on partial products at the bit level when generated in pairs, systematically exposing redundancy across output rows to reduce the number of PP generators and selectors. This redundancy is further exploited using a lightweight correction unit that derives a row's output directly from its paired row. Additionally, for block-circulant MVMs, relocating the shift-accumulate and correction units of each sub-MVM further reduces resource usage. The compressed network achieves weight compression of up to 93.6% while maintaining accuracy on the TIMIT and LibriSpeech-100h benchmarks. On a field-programmable gate array, COBALT achieves superior overall efficiency relative to state-of-the-art LSTM accelerators.

In this paper, we present a novel framework for recovering individual footprints and pedestrian spatial behaviors in dense crowds. This is a highly challenging problem, as pedestrians can merge or block one another often and for prolonged periods in crowded areas, leading to strongly entangled observations. Our approach jointly leverages learned crowd semantics and principled reasoning over persistent radar observability loss to systematically infer individual spatial footprints. At the macroscopic level, we learn spatial usage patterns of the crowd from radar point clouds, while at the microscopic level we develop a physics-informed model of radar observability loss caused by sustained merging and occlusion. Building on this macro-micro structure, we introduce a semantic-guided multi-hypothesis reasoning foundation that evaluates competing hypotheses via principled reasoning over many-to-one and one-to-none observation mappings, enabling untangling of individual spatial footprints. We evaluate the framework through 18 real-world experiments using an off-the-shelf mmWave radar, involving crowds of up to (and including) 10 individuals across six categories of diverse pedestrian behaviors, and four environments. Our methodology robustly recovers individual spatial footprints, even during prolonged and persistent observability loss, demonstrating strong structural consistency with the ground-truth. The recovered footprints further enable accurate zonal analysis of pedestrian space usage. Finally, we present an ablation study and further compare against state-of-the-art. The proposed foundation substantially outperforms the state-of-the-art across all experiments, achieving an average median DTW of 30.5 cm compared to 95.94 cm, highlighting the framework's robustness to frequent and prolonged merging and occlusion.

Distributed target tracking is essential for scalable and robust sensing systems, as it enables multiple radar nodes to cooperatively estimate a target state without relying on a centralized fusion center. In this paper, we present a fully distributed framework for single-target tracking in frequency-modulated continuous-wave (FMCW) radar networks. Each monostatic radar node observes local range and Doppler measurements and exchanges information only with neighboring nodes. Two consensus optimization-based estimators are developed. First, a Distributed Maximum A Posteriori (D-MAP) estimator is formulated for batch-based tracking, where prior state information is incorporated and the resulting optimization problem is solved using consensus-based alternating direction method of multipliers (ADMM). Second, a Distributed Extended Kalman Filter (D-EKF) is proposed for recursive tracking, where each node performs local prediction and correction followed by consensus ADMM-based information exchange. We derive the posterior Cramér-Rao lower bound (PCRLB) as a theoretical performance benchmark. Our simulation results show that D-MAP improves the accuracy of the estimation over the distributed maximum-likelihood baseline. These results demonstrate that the proposed framework provides a scalable and robust alternative to centralized radar tracking, given only local inter-node communication.

Deep reinforcement learning (DRL) has recently gained a great attention due to its real-time adaptation and effectiveness in complex optimization problems. This paper investigates the optimal deployment of millimeter-wave (mmWave) base stations (BSs) in a realistic, non-convex campus topology. The optimization problem is NP-hard, due to the non-convex, non-smooth nature of the max-min fairness objective. To overcome these constraints, we formulate the BS placement as a Markov Decision Process (MDP) and systematically benchmark four DRL schemes: a discrete single-agent Deep Q-Network (DQN), a spatially partitioned Multi-Agent DQN, a continuous single-agent Deep Deterministic Policy Gradient (DDPG), and a geographically partitioned multi-agent DDPG framework. Numerical evaluations reveal that the multi-agent DDPG approach substantially outperforms single-agent in dense scenarios. Additionally full coverage is achieved, and a fairness Jain's index of 0.94 is obtained. Finally, the multi-agent demonstrates highly efficient computational convergence of dense scenarios with $400$ users.

Implantable neural interfaces require low-power real-time signal processing to remain within strict thermal and bandwidth constraints, motivating lightweight feature extraction methods for on-chip spike sorting. This work presents the Walsh-Hadamard Transform (WHT) as a hardware-efficient feature extraction method for neural spike classification. WHT can be implemented using only adders, subtractors, and registers without coefficient memory. WHT performance is compared against the Compressed Hadamard Transform (CHT) and Principal Component Analysis (PCA), improving mean F1-scores from 55-60% to 70-75% on difficult high-noise datasets and from 90-95% to 95-99% on all other simulated datasets. In addition to improved classification performance, WHT demonstrates greater robustness to noise, downsampling, reduced training size, and distance metric selection, maintaining standard deviations typically below 5%, while CHT and PCA reach up to 10% under high-noise conditions.

Integrated sensing and communication (ISAC) enables the tracking and detection of passive objects, including the human body, by leveraging standard wireless communication signals. While already standardized in IEEE 802.11bf for Wi-Fi, ISAC is currently being defined by 3GPP for the upcoming generations of cellular networks. This transition scales the ISAC technology from localized, uncoordinated Wi-Fi deployments into a pervasive, centralized network deployment managed by mobile network operators. However, such nationwide coverage introduces critical user privacy challenges that must comply with stringent data protection legislative frameworks, most notably the General Data Protection Regulation in the European Union. This paper provides a comprehensive analysis of applicable norms to cellular ISAC, highlights the open technical challenges in its implementation, and explores potential mitigation strategies at both the architectural and signal processing levels, establishing tiered data access levels for various stakeholder categories and detailing how these protocols can be lawfully integrated into cellular network architectures.

This paper studies integrated sensing and communications (ISAC) over a hybrid system that seamlessly combines legacy cellular base stations with distributed cell-free (CF) access points (APs). We propose a hierarchical ISAC architecture where a central base station (CBS) serves its near users and simultaneously operates as a monostatic radar for aerial target detection, while distributed APs—many idle under user-centric clustering—act as cost-free bistatic receivers. The CBS jointly handles communication processing and multi-static sensing fusion, reducing fronthaul overhead compared to conventional cell-free ISAC. To achieve this, a five-phase time-division duplexing workflow with precise ISAC role assignment is specified. Closed-form expressions for spectral efficiency and multi-static sensing signal-to-noise ratio analytically characterize the communications–sensing Pareto frontier. Numerical results confirm that the proposed hierarchical design simultaneously achieves higher sum throughput and superior sensing accuracy than conventional cell-free ISAC.

In this work, we show that pilot contamination can be exploited to jointly increase the uplink sum rate and reduce pilot overhead. By leveraging the effect of pilot contamination on minimum mean-square-error (MMSE) channel estimates, we demonstrate that allowing users to share or use correlated pilots can, in certain regimes, yield better sum rate than with mutually orthogonal pilots or perfect channel state information (CSI) under maximum-ratio (MR) combining. To establish this result, we derive the use-and-then-forget (UATF) capacity bound for an arbitrary pilot design under correlated Rayleigh fading with MR combining. We show that, in the presence of pilot contamination, the key mechanism is the directional suppression of the channel estimate towards nearby interfering angular regions, enabling interference reduction. Numerical results confirm this behavior, demonstrating the sum-rate gain and showing that it grows with the number of antennas.

Sea ice drift underpins air-sea-ice coupling and model evaluation, but the Southern Ocean remains chronically undersampled relative to the Arctic. This work presents a cross-polar comparison of GNSS-tracked ice-drift time series using rotary spectral analysis, principal component analysis, inter-buoy coherence, complementary shape- and amplitude-sensitive spectral differences. Arctic data are aggregated into Beaufort Gyre and Transpolar Drift seasonal composites (2017–2024); Antarctic data comprise eight heterogeneous campaigns (2000–2022) with small buoy counts and short durations. After common quality control, the clearest cross-polar similarity is spectral organisation: both regions show strong low-frequency variance and a hemisphere-appropriate rotary enhancement near the Coriolis frequency, interpreted cautiously as a combined inertial–semidiurnal response. Coherence is highest at synoptic scales and declines toward higher frequencies. Southern Ocean campaigns occupy a higher-energy envelope, so similar band structure does not imply similar drift amplitude. Principal components capture much array-scale motion, but high cumulative variance does not substitute for dense sampling of smaller-scale processes. The results provide transferable frequency-resolved benchmarks (dominant variance below 0.5 cpd and in the near-inertial band, an expanded Arctic reference, and explicit separation of spectral shape from amplitude) to inform future observing-system design without assuming identical dynamical regimes.

Egocentric vision systems capture human behavior from visible cues, but overlook physiological indicators of autonomic states such as stress, engagement, and attention. Heart rate variability (HRV) is a widely used noninvasive marker of autonomic regulation under stress. HRV reflects small timing differences between successive heartbeats and has so far been out of reach for egocentric platforms, where motion and noise in gaze video mask exactly this fine-grained timing. We propose EgoHRV, a method that estimates HRV as well as heart rate (HR) from the gaze cameras that are already integrated into egocentric headsets. Our pipeline combines a 3D backbone with a novel low–high decomposition module that extracts the blood volume pulse (BVP) signal from gaze video. Our cross-domain pretraining aligns the frequency-domain representations of contact-based and camera-derived signals. This alignment gives EgoHRV the temporal precision to recover HRV from the subtle fluctuations in gaze video. EgoHRV achieves state-of-the-art accuracy for HR and HRV estimation from egocentric video, and its uncertainty-aware design improves downstream behavioral modeling. Integrating our HRV estimates and confidence measures into EgoExo4D's proficiency estimator raises accuracy by 17.8%. Beyond skill, continuous HRV estimation also opens egocentric systems to stress- and arousal-aware estimation tasks. Code: https://github.com/eth-siplab/EgoHRV

Non-invasive continuous blood pressure (BP) monitoring using photoplethysmography (PPG) is a promising alternative to cuff-based measurements. However, existing PPG-based BP estimation studies predominantly rely on aggregated performance metrics (e.g., mean absolute error) computed over entire evaluation intervals, which can obscure model failures during rapid BP fluctuations and limit clinical relevance. In this work, we propose a fluctuation-aware evaluation framework for PPG-based BP estimation based on time-series change point detection. Instead of heuristic BP thresholding (e.g., $\Delta\mathrm{BP} > 10\mathrm{mmHg}$), we identify BP change points by capturing abrupt distributional shifts in BP trajectories and evaluate estimation performance specifically during these fluctuation periods. Our analysis shows that several state-of-the-art models exhibit substantial performance degradation around BP change points, and that periodic test-time calibration is insufficient to handle such dynamic BP variations. To address this limitation, we introduce a targeted re-calibration framework triggered by detected BP change points, improving robustness without modifying model architectures. To the best of our knowledge, this is the first systematic evaluation of PPG-based BP estimation from a BP change point perspective, highlighting the importance of fluctuation-aware evaluation and calibration for real-world continuous BP monitoring.

High-precision Global Navigation Satellite System (GNSS) services rely on accurate orbit, clock, atmospheric, and hardware-bias corrections generated from reference observations. These products are traditionally derived from terrestrial reference networks, whose performance strongly depends on the density and geographic distribution of ground stations. Consequently, sparse or regionally concentrated networks can suffer from tracking gaps and limited global observability, reducing their ability to support globally consistent high-precision products. Low Earth Orbit (LEO) constellations equipped with onboard GNSS receivers and inter-satellite links (ISLs) can serve as a network of spaceborne reference stations, offering a promising way to extend terrestrial reference networks into space, thereby improving observability for ground networks and enabling future direct correction broadcast. However, most existing network-based GNSS correction-generation workflows assume that observations can be centrally collected and processed. This assumption fails in practical ground–to–space architectures, where dynamic satellite geometry and system constraints render communication links intermittent, asymmetric, capacity-limited, and lossy. To address these limitations, this paper proposes a decentralized processing architecture that coordinates the ground-to-space GNSS reference network. By modeling ground stations and LEO satellites as interacting subnetworks over a dynamic graph, our approach allows frequent intra-tier communication while restricting cross-tier exchanges to compact estimation summaries transmitted opportunistically under probabilistic link availability....

Distributed detection of an unknown deterministic signal is studied in a wireless sensor network with low-cost nodes. Sensors apply one-bit quantization to noisy observations and transmit over Rayleigh fading to a Fusion Center (FC). Score tests, including variants using observed Fisher information, are proposed as low-complexity alternatives to the generalized likelihood ratio test (GLRT). The FC performs joint decoding and fusion, optimizing quantizers in a channel-aware manner to improve asymptotic detection performance. Simulations confirm the appeal of score-based tests under realistic wireless conditions.

The idea of using Lipschitz extensions [1,2], or Gradually Varied Functions (GVFs)[3], for mesh-free scattered data reconstruction was proposed by the author in 2012 [4]. However, its practical application to modern mesh-free reconstruction methods has not been fully explored. Motivated by recent advances in computational tools, including AI-assisted mathematical programming and software development, we revisit this idea and investigate the use of a Lipschitz extension as an initialization step for Moving Least Squares (MLS) reconstruction [5,6]. Our computational experiments indicate that this initialization significantly improves the stability and reconstruction accuracy of MLS under sparse and irregular sampling. This is a preliminary study intended to establish feasibility; a fuller evaluation with additional benchmarks and comparisons is left to future work.

Automatic Modulation Recognition (AMR) plays a crucial role in enabling robust, adaptive, and secure communication for military and civilian applications. Deep learning has enabled effective AMR methods that overcome the computational inefficiency of traditional approaches. However, these deep learning based methods often degrade significantly in low SNR conditions, where noise obscures modulation-discriminative waveform features. In this paper, we propose Bounded Residual Feature-wise Linear Modulation (BR-FiLM), a channel-quality conditioning block for AMR, which can be inserted into AMR classifiers with intermediate feature representations. We construct BR-FiLMNet by inserting the proposed BR-FiLM block into a Multi-Channel Convolutional Long Short-Term Deep Neural Network (MCLDNN) backbone. BR-FiLMNet conditions convolutional, recurrent, and dense features through gated residual corrections while preserving the original I/Q-driven feature path. Experimental results on RadioML 2016.10a show that BR-FiLMNet improves mean accuracy from 61.79% to 67.74% and low-SNR accuracy (SNR <= 0 dB) from 37.12% to 46.44% compared to MCLDNN. We further evaluate BR-FiLMNet against recent transformer-style baselines. The results indicate that BR-FiLMNet delivers significant and reliable performance gains, as validated through paired statistical testing.

Classical seismic data reconstruction relies on manually designed structural priors and iterative operators, whose coupled design space is far larger than manual trial and error can explore systematically. Deep-learning methods encode the reconstruction rules in learned weights rather than in an explicit operator that can be inspected and modified. We propose SeisEvo (Seismic Algorithm Evolution), which does not optimize a single reconstruction result but searches for the algorithm that produces it. Starting from a classical reconstruction algorithm, an LLM-driven multi-agent search modifies only the components that the user has opened for editing, without prescribing the mechanism to be discovered. Candidates that violate the physical constraints of the task are rejected outright, and the remaining ones are scored by execution. The output is neither an agent system nor a neural network, but a standalone white-box algorithm that requires no agent or neural network at inference time. For interpolation without added noise, the search discovered a residual-gated, phase-aligned dip-consistency projection; Evo-POCS improves the SNR over classic POCS by 3.49 dB on average across missing ratios from 30% to 70%. For simultaneous interpolation and denoising, it discovered a reliability-grouped singular-value shrinkage; Evo-MSSA improves the average reconstruction SNR by more than 7 dB over classic MSSA and by more than 3 dB over a stronger rank-reduction baseline. Both operators retain their gains on data not used during the search. To the best of our knowledge, this is the first study to formulate the design of a seismic reconstruction operator as a constrained, LLM-driven program evolution task. Agentic algorithm evolution can thus complement deep learning in discovering explicit, inspectable, and deployable seismic processing algorithms.

Reliable quantitative analysis of dynamic contrast-enhanced MRI requires high-quality spatiotemporal reconstructions at high undersampling rates. Scan-specific reconstructions using Gaussian and Gabor primitives have shown promising results without the need for large training datasets, but have not addressed the additional dimension of dynamic contrast. We propose a multi-dimensional, primitive based framework for dynamic contrast-enhanced MRI reconstruction that disentangles the underlying anatomy, the dynamic contrast enhancement, and residual motion into separate temporal basis functions, thereby enabling a geometrical interpretation of the representation. We show that this architecture achieves performance competitive with conventional reconstruction methods, both in reconstruction quality and in the accuracy of extracted aorta and kidney enhancement curves. The modular tier design extends naturally to additional dynamic factors and higher acceleration rates. Code available at https://github.com/compai-lab/2026-GaborDCE-spieker.

State estimation is commonly posed in either a probabilistic or an unknown-but-bounded framework. The former requires a fully specified noise distribution, typically with unbounded support, while the latter yields guaranteed enclosures that carry no probabilistic weighting. Bridging these noise descriptions, this paper proposes a zonotopic mixture noise model, in which the noise is generated by drawing a zonotope from a finite collection according to fixed probabilities and then realizing an arbitrary element of it. For this noise model, we derive the zonotopic mixture filter, which propagates a bank of zonotopic Kalman filters over mode histories, discards the histories falsified by the data, and weights the surviving ones by their relative probability. The resulting state enclosures yield guaranteed coverage probabilities and remain valid for every noise realization compatible with the bounds, and a greedy mixture reduction scheme preserves these statistical guarantees while keeping the representation tractable. Numerical examples illustrate the proposed approach and its potential benefits over related state estimation methods.

Embodied intelligence is shifting artificial intelligence from passive digital perception toward active physical interaction. However, foundation-model-enabled embodied agents face a fundamental tension between open-world cognition and resource-constrained deployment. On-device models are limited by computation, memory, and energy budgets, whereas cloud-centric solutions introduce latency and reliability risks over dynamic wireless links. Edge general intelligence provides a promising cognitive backbone, but existing frameworks still lack physical grounding, action awareness, and mechanisms for actively acquiring useful physical experience. To address these limitations, this article introduces edge-native embodied intelligence (ENEI), an action-aware wireless edge framework that integrates embodied agents, the 6G communication and networking fabric, and edge cognitive services into a 6G-mediated bidirectional edge-embodiment loop. Along the edge-to-embodiment axis, confidence-aware assistance and edge-driven generative adaptation enhance local autonomy under out-of-distribution (OOD) conditions. Along the embodiment-to-edge axis, value-of-experience guided active embodied federated learning enables physical actions to generate informative experience for continuous edge model evolution. The 6G fabric supports both directions through goal-oriented transmission and programmable radio-resource allocation. Two case studies on OOD drone navigation and mobility-driven federated learning illustrate the feasibility and communication efficiency of the proposed mechanisms. ENEI provides a unified perspective in which edge cognition strengthens embodied action, while embodied agency actively enriches edge cognition, laying the foundation for scalable, adaptive, and self-evolving embodied wireless systems.

The integration of sensing, communication, and intelligence is becoming a key enabler for sixth generation (6G) wireless systems, where intelligent terminals are expected to simultaneously support efficient link establishment and reliable environmental sensing. However, existing studies mainly exploit sensing information or communication information to address a single task, such as channel prediction or environment reconstruction. Motivated by the shared dependence of optical and radio-frequency signals on the surrounding environment, we propose the electromagnetic world model (EMWM), the first unified framework for joint environment reconstruction and channel prediction. EMWM learns a common electromagnetic representation with the potential to provide a modeling foundation for 6G tasks. Specifically, partial channel state information (CSI) and multi-view red-green-blue (RGB) images are encoded into CSI and visual tokens and jointly processed by a hierarchical world-model backbone with local and global aggregation. Based on the learned representation, a mixture-of-experts (MoE)-based CSI prediction head reconstructs the complete CSI, while a depth prediction head estimates multi-view depth maps that are further converted into three-dimensional (3D) point clouds. Moreover, a large-scale multi-modal dataset is constructed based on a campus digital twin. Experimental results show that EMWM outperforms conventional neural network and large language model (LLM) baselines in both CSI prediction and environment reconstruction, achieving a squared generalized cosine similarity (SGCS) of 0.9699 for CSI prediction while demonstrating robustness across different signal-to-noise ratio (SNR) conditions and zero-shot generalization at 28 GHz.

Intelligent channel state information (CSI) feedback is essential for realizing the high capacity and spectral efficiency goals of future 6G systems, yet existing deep learning solutions face a trade-off between model generalization and scenario-specific performance. Large neural networks generalize well but incur high computational and tuning costs, while small models excel in particular environments but require repetitive costly end-to-end training for each base station (BS). To address these challenges, we introduce a model repository-based deployment framework in which a centralized AI data center maintains a catalog of scene-specific CSI models. The repository is enhanced with a Learnware-based framework, where each model is associated with a specification including semantic part (network architecture parameters) and statistical part (codeboo-fingerprint embeddings of training-data distributions). A BS submits only its local statistical specifications to retrieve the most relevant pre-trained model, enhancing data privacy by avoiding raw CSI transmission and drastically reducing retrieval latency and communication overhead. We further develop a data-driven search strategy that matches codebook fingerprints to model performance, achieving over 90% selection accuracy. In simulations, our scheme yields 18.8% and 57.7% performance improvements over the General Model in LOS and NLOS scenarios, respectively while reducing local fine-tuning by up to 1000 samples and 100 epochs. This Learnware-based approach minimizes redundant training, maximizes model reuse, and supports rapid,privacy-enhancing deployment of CSI feedback models.

The quantum approximate optimization algorithm (QAOA) targets NP-hard maximum-likelihood (ML) detection in multiple-input multiple-output (MIMO) systems. Existing $M$-ary quadrature amplitude modulation (M-QAM) detectors design angles by expected Ising energy: online per instance, warm-started, or ramped, while train-once designs remain B/QPSK-only or block-local, leaving M-QAM without a size-scalable benchmark. Their infinite-size spin-glass theory assumes independent disorder, matching the retained covariances at B/QPSK but not M-QAM's correlated couplings and fields. We develop a correlated infinite-size multi-species spin-glass framework whose covariance-matched evaluators make that energy an offline objective with a size-scalable benchmark. In addition, the ML rate, the exponential rate of sampling the ML string, is for the first time exploited for QAOA angle design in MIMO detection. The energy evaluator is $q$-free at $O(p\,4^p)$ cost while the ML rate transfers angles from a fixed $q_{\rm ref}$-qubit reference. Tests reach 4096-QAM, 128 antennas, $p=30$ and per-symbol SNR 0-45 dB. In simulations, ML rates fall as a power law $r_0\,p^{-\alpha}$, with larger exponents for the sampling design, which tracks exact ML at $5\times5$ 16-QAM (0-20 dB) and $3\times3$ 64-QAM (8-28 dB) while its bit-error rate (BER) advantage widens with SNR to two orders of magnitude. The approach points toward near-optimum decoding on deeper noiseless fault-tolerant quantum (FTQ) circuits.