To address UAV obstacle avoidance under motion blur, we propose Feature-Domain Motion Blur Decoupling Planner (MBD-Planner), a real-time feature-domain motion blur decoupling framework that jointly disentangles blurred visual features and optimizes trajectories in an end-to-end manner. Unlike methods relying on explicit 3D reconstruction or ignoring blur in planning, MBD-Planner introduces a lightweight dewarping module that transforms blurred features into virtual sharp ones via depth back-projection and pose compensation, achieving sub-millisecond latency. A blur-aware optimization objective further couples camera motion and trajectory smoothing, while privileged training guided by ESDF gradients enables robust learning without expert demonstrations. Experimental results demonstrate that MBD-Planner enables reliable, smooth, and real-time navigation in visually degraded environments, effectively bridging the gap between theoretical planning and robust navigation.

Cell-Free massive MIMO (CF mMIMO), as a cornerstone for future wireless communication networks, can significantly enhance spectral efficiency, reduce energy consumption, and achieve seamless coverage. This paper investigates dynamic resource allocation in CF mMIMO systems provisioning integrated sensing and communication (ISAC) and involving multiple users and heterogeneous services. To address the rapidly varying channels, high complexity, task-dependent relevance and insufficient information sharing among access points (APs) arising from ISAC requirements, we propose a novel Hierarchical Attention-Driven Multi-Agent Reinforcement Learning (HADMARL) framework. A higher-layer “Commander Controller” outputs the AP-user pairing matrix, which is passed to a lowerlayer “Worker Controller” that leverages the central attention mechanism to perform fine-grained resource allocation. Simulations demonstrate that HAD-MARL achieves about a 10% improvement in system utility score compared to benchmarks, with the performance gains becoming more pronounced as the number of APs and users increases. We empirically demonstrate and visualize the effect of the central attention mechanism on the formation of efficient collaboration among APs. The source code is available at: https://github.com/Andyyy2000/HAD.git.

Magnetic induction communication (MIC) is a promising technology for through-the-earth (TTE) communication. Previous studies on the MIC range have often overlooked the impact of eddy losses caused by underground materials. For TTE MIC, significant eddy losses complicate the analysis of the effective MIC range, which is vital for optimizing performance but has never been addressed in the literature. Accounting for the conductivity and permittivity of the underground medium, this paper derives the effective MIC range in TTE MIC, along with a closed-from expression that predicts the optimal carrier frequency to maximize this range. Finite element simulations validate the analysis, demonstrating that the optimal carrier frequency can significantly enhance the MIC range. It is also revealed that optimizing the antenna radius is effective in extending the MIC range for TTE and vehicle MIC applications.

The emergence of movable antenna (MA) technology provides a promising way to enhance wireless sensing and communication by introducing spatial degrees of freedom through dynamic array reconfiguration. In near-field localization, achieving high resolution at low cost necessitates the adoption of sparse arrays. However, such sparsity tends to introduce spatial ambiguity due to aliasing effects. To resolve this resolution-ambiguity dilemma, this paper proposes an MA-enabled array zooming (AZ) system. First, we design a multi-measurement array zooming system that dynamically adjusts antenna spacings. By fusing the observational information from different measurements, the proposed AZ system effectively mitigates spatial aliasing while maintaining spatial resolution. Second, to quantify the performance limits under the severe multi-modal distributions inherent in sparse near-field sensing, we theoretically analyze the false peak distribution and derive a tighter performance lower bound, which incorporates the false detection probability. Third, considering that multiple false peaks may exist in practical multi-modal distributions, we propose an optimization algorithm for the AZ system to suppress false peaks and minimize the localization error. Extensive numerical results demonstrate that the proposed AZ strategy adaptively optimizes array configurations under varying signal-to-noise ratios (SNRs), substantially outperforming both conventional fixed-spacing arrays and Cramer-Rao bound (CRB)-based AZ benchmarks in localization accuracy.

Extended reality (XR) is becoming a crucial application in the shift from 5G to 6G networks, providing immersive experiences that combine real and virtual environments. However, the real-time transmission of XR video encounters challenges with high data rates and low latency, particularly due to its frame-by-frame delivery, which is highly sensitive to network fluctuations. To address this challenge, we propose a comprehensive framework for the collaborative transmission of real-time XR video and haptic feedback that emphasizes joint source-channel optimization (JSCO). Specifically, this framework involves two critical components: adaptive video quality selection at the source and efficient wireless resource allocation over the channel. By optimizing both the video content and the underlying communication resources, we aim to enhance the overall transmission efficiency and user satisfaction. We also propose a dual-layer deep reinforcement learning (DRL) algorithm for solving JSCO problem, termed as proximal policy optimization combined with multi-agent Transformer (PPO-MAT), which operates within a hierarchical framework to effectively tackle the issue of inconsistent time scales. Through extensive experiments, we demonstrate that our proposed method significantly outperforms existing baseline solutions, achieving at least a 14.39% enhancement in quality of experience (QoE) and a 67.71% increase in the number of satisfied users. These results underscore the effectiveness of our joint optimization approach, ultimately paving the way for more robust and responsive XR applications in dynamic network environments.

Decentralized Federated Learning (DFL) enables collaborative model training without central coordination. However, DFL faces challenges in dynamic networks, where existing methods struggle to balance consensus rate and communication efficiency, while overlooking practical issues such as topology variation. This paper presents Dynamic AirComp-enabled DFL (DA-DFL), a novel framework that integrates over-the-air computation (AirComp) with the BASE-GRAPH consensus algorithm for efficient DFL over dynamic topologies. The convergence analysis for DA-DFL under dynamic settings is conducted to reveal the influence of the consensus period and communication errors. We define communication overhead metrics, and jointly optimize transceiver beamformers and dynamic topologies. A topology matching algorithm is developed to reduce communication overhead by aligning logical and physical topologies. Experiments show significant gains of DA-DFL in communication efficiency, e.g., reducing communication links and distances by up to 42% and 50%, respectively, compared to benchmarks.

In reconfigurable intelligent surface (RIS)-aided integrated sensing and communication (ISAC) systems, beamforming design based on minimizing the Cram & eacute;r-Rao bound (CRB) for target direction-of-arrival (DoA) estimation is pivotal for sensing capability enhancement. However, due to practical hardware limitations, phase-shift errors (PSEs) exist at the RIS reflectors and cause performance deterioration. To reduce the adverse impact of PSEs, we develop a novel stochastic optimization (SO)-based expected CRB (ECRB) minimization framework, where the ECRB is defined as the expectation of CRB taken over random PSEs, statistical channel state information (CSI), and historical DoA estimates following known prior distributions. Specifically, we formulate an SO problem to minimize the ECRB, subject to an ergodic achievable sum-rate (EASR) constraint. To solve this non-convex problem, we propose a novel penalty-based block stochastic gradient descent (PBSGD) method. In this method, we first introduce a penalty factor to move the EASR constraint into the objective function. The optimal penalty factor is rigorously proved to be determinable via a bisection search. Then, we design a projected block stochastic gradient descent process to update transmit beamformers and RIS phase shifts alternately with guaranteed convergence. Simulation results demonstrate that our proposed method outperforms several benchmarks, including random phase-shift design, sensing-only beamforming, and state-of-the-art semidefinite relaxation (SDR)-based CRB optimization techniques, while exhibiting enhanced robustness against PSEs.

This paper studies the near-field localization problem under dynamic scenarios, which harnesses the non-line-of-sight (NLoS) components, in a reconfigurable intelligent surface (RIS)-aided system equipped with extremely large-scale multi-input multi-output (XL-MIMO). To reduce the complexity of the position estimation, the subarray far-field model is employed to approximate the near-field channel. A factor graph within a Bayesian framework is constructed to detail the probability transition relationship among the relevant variables. Based on the message passing in this factor graph, a near-field localization algorithm is developed to estimate the marginal probability distributions of the UE's and scatterers' positions in each time slot. The misspecified Cram & eacute;r-Rao Lower Bound (MCRLB) is derived to evaluate the performance of the algorithm under the subarray far-field model. To explore the localization potential of the system, a closed-form solution for a low-complexity directional beamforming design and a robust beamforming design based on the gradient descent method (GDM) are further proposed. Numerical results demonstrate that the proposed algorithm outperforms the benchmark schemes, and validate the performance gain of harnessing the NLoS components.

Time-of-arrival (ToA)-based user localization typically requires precise clock synchronization between base stations and user equipments, making the localization and synchronization problems tightly coupled with each other. This paper considers a distributed multi-input multi-output (MIMO) orthogonal frequency-division multiplexing (OFDM) system and addresses joint multi-user localization and clock synchronization within an integrated sensing and communication (ISAC) framework. Unlike existing works that only consider clock bias, we account for the impacts of both clock bias and clock skew on MIMO-OFDM signals. Specifically, clock bias introduces a constant offset in path delays, whereas clock skew causes a mismatch in the OFDM symbol durations between the transmitter and the receiver, resulting in linearly varying delays across OFDM symbols. We formulate the joint localization and synchronization problem within a Bayesian framework. Based on variational message passing and the sum-product rule, we propose a message passing algorithm, termed Bayesian Localization and Clock Synchronization (BLACS), which jointly estimates the positions, clock parameters, and velocities of multiple users. Simulation results show that accounting for clock skew significantly improves localization accuracy compared to the baseline methods. Moreover, the proposed BLACS algorithm achieves performance close to the Bayesian Cram & eacute;r-Rao Bound, demonstrating its effectiveness and near-optimality.

Beamforming design for extremely large-scale multiple-input multiple-output (XL-MIMO) systems is challenging due to prohibitive computational complexity and complex near-field propagation effects. To address this, this paper introduces a holographic beamforming paradigm that reformulates the design from optimizing variables at spatially discrete antenna locations to shaping a continuous electromagnetic wave function over the array aperture, effectively mitigating the growth of algorithmic complexity as the array scale increases. We apply this paradigm to the challenging dual near-field (DNF) scenario, where strong transceiver coupling severely degrades conventional iterative algorithms. In this case, we propose a novel Virtual Point Source (VPS) method, which approximates the ideal wave function with a single and analytically tractable spherical-wave. A rigorous geometric-optical analysis is provided to show that the optimal VPS location can be determined in a fully non-iterative manner, thus decoupling the coupled DNF problem. The proposed method is demonstrated in an intelligent reflecting surfaces (IRS)-assisted system, where simulation results show that our non-iterative approach achieves performance comparable to converged alternating-optimization (AO) algorithms, while incurring significantly lower complexity and avoiding convergence uncertainty. This work offers a new theoretical framework for holographic beamforming design in XL-MIMO systems.

Magnetic induction (MI) communication (MIC) has emerged as a promising candidate for underground communication networks due to its excellent penetration capabilities. Integration with Space-Air-Ground-Underground (SAGUI) networks in next-generation mobile communication systems requires a well-defined network architecture. A recent discovery in MIC research, MI fast fading, remains in its early stages and presents unique challenges. This paper provides a comprehensive survey on through-the-earth (TTE) MIC, covering MI applications, channel modeling, point-to-point MIC design, relay techniques, network frameworks, and emerging technologies. We compare various MIC applications to highlight TTE-specific challenges and review the principles of channel modeling, addressing both MI slow fading and MI fast fading, along with its potential impact on existing MIC theories. We conduct a fine-grained decomposition of MI channel power gain into four distinct physical parameters, and propose a novel geometric model to analyze MI fast fading. We also summarize MI relay techniques, examine crosstalk effects in relay and high-density networks, and explore key research tasks within the OSI framework for a holistic MI network protocol in SAGUI. To bridge the gaps identified, we propose a MIC framework that supports TCP/IP and Linux, enabling full implementation of existing and emerging MIC solutions. This framework empowers researchers to leverage Linux resources and deep learning platforms for accelerated development of MIC in SAGUI networks. Remaining research challenges, open issues, and promising novel techniques are further identified to advance MIC research.

Intelligent reflecting surface (IRS)-assisted integrated sensing and communications (ISAC) systems have been extensively studied to meet higher sensing requirements. For detection-oriented IRS-assisted ISAC problems, most studies have over-looked the detection interference caused by clutters and modeled simplified point-like targets. This paper investigates extended target detection in IRS-assisted ISAC systems within clutters. We present an optimal generalized likelihood ratio test detector and derive the corresponding probability of detection (PD) and probability of false alarm in closed form. Then, we jointly optimize the active and passive beamforming of the base station and IRS to maximize the PD under multi-user equipment (UE) communication rate constraints and the total transmit power constraint. We first simplify the complex objective function by proving the invariant property of a subspace projection matrix. We then present a novel alternating optimization (AO)-based algorithm to decouple the original problem into two subproblems, consequently convexified and solved using the semidefinite relaxation method. Simulations demonstrate the convergence of the proposed algorithm. The PD performance and the communication and sensing trade-off are significantly improved, compared to benchmarks.

Intelligent reflecting surface (IRS) is a promising technology, yet its physical design, particularly the element spacing, generally relies on the conventional half-wavelength ( lambda/2 ) rule from active arrays. This letter investigates the validity of this rule for IRS-aided systems with two-hop cascaded channels. In this letter, we first derive the spacing requirements from first principles by applying the Nyquist sampling theorem to both the reflection-angle and the cascaded-channel-angle domains. We then distinguish between Visible Grating Lobes (VGLs) and a more subtle phenomenon, termed Ghost Grating Lobes (GGLs). Our analysis confirms that while lambda/2 spacing is indeed the threshold to prevent VGLs, whereas a stricter quarter-wavelength ( lambda/4 ) spacing is the actual threshold required to avoid GGLs due to the doubled angular bandwidth of the cascaded channel. Simulation results are presented to validate our theoretical findings. It is shown that element spacing between lambda/4 and lambda/2 leads to significant channel correlation and system-level interference, even in the absence of VGLs. These results establish a refined two-threshold framework that provides a more complete guideline for IRS design, especially in interference-limited systems such as cell-free Multiple-Input Multiple-Output (MIMO) systems.

Timing uncertainty in containerized wireless industrial control arises from the combined effects of wireless communication, protocol processing, and container virtualization. This letter proposes a reinforcement learning (RL)-assisted proportional–integral–derivative (PID)-compatible compensation framework for such uncertainty. A nonlinear controller-level model represents bounded delay and correlated read/write losses while distinguishing candidate from plant-applied inputs. A closed-form, safety-projected additive compensator retains the PID structure, while a cloud-trained RL policy schedules the controller parameters. Sufficient conditions are derived for uniform ultimate boundedness of the tracking error on a compact validation region under intermittent communication failures. Quadruple-tank results show that the proposed method achieves the lowest mean squared error, integral absolute error, and worst-case settling time among all evaluated methods, including fixed and robust controllers.

Autonomous aerial exploration in dynamic industrial environments requires global coverage efficiency and local safety under moving obstacles. Existing planners usually rely on static or quasi-static space decomposition and therefore exhibit oscillatory motion, frequent replanning, and poor mission continuity when workers or vehicles temporarily block topological passages. This article presents spatio-temporal aerial guidance for exploration (STAGE), a hierarchical spatio-temporal framework for coverage-path-guided aerial exploration. The key idea is to model graph traversability as a departure-time-conditioned quantity rather than a static geometric adjacency. Specifically, we construct an edgewise 4-D cost matrix that embeds predicted obstacle occupancy and a bounded waiting strategy, solve a time-dependent asymmetric traveling salesman problem with an Lin-Kernighan-Helsgaun (LKH)-based heuristic to obtain a temporally consistent global visitation order, and refine the local execution via a sequential ordering problem under kinodynamic constraints. We further clarify the approximation scope of the hierarchical solver, analyze the bounded-error property of the waiting search, and discuss the risk-budget calibration used on the uncrewed aerial vehicle platform. Simulations and field experiments on an industrial quadrotor show up to 7.8% reduction in exploration time and 54.8% reduction in mean squared jerk relative to static or quasi-static baselines. These results indicate that STAGE improves mission continuity and edge-deployable decision making for industrial inspection in dynamic workspaces.

This paper investigates a challenging maritime sensing security issue, where a target unmanned aerial vehicle (UAV) intends to evades multiple unauthorized radars detection with the help of a decoy UAV. For the considered maritime wireless sensing, the radars are installed on the moving ships and their accurate positions are imperfectly known. Moreover, the sea surface clutter and platform-induced Doppler effects need to be considered. To address these challenges, we propose a dual UAV-mounted reconfigurable intelligent surface (RIS) enabled strategy, where RIS 1 is deployed on the target UAV for directing all probing signals toward the decoy UAV to shield the actual target, while simultaneously RIS 2 mounted on the decoy UAV can redirect the reflected signals back to the adversary radars directions to deceive the radars detection. Furthermore, based on the sensing thresholds and estimation errors, the target UAV can also intelligently adjust its flight trajectory to move closer to/away from the decoy UAV/unauthorized radars to obtain a better system performance. In such a setup, the maximization of the received echo signal powers from the direction of the decoy UAV is formulated into an optimization problem while suppressing that from the direction of the target UAV below a certain threshold, by jointly designing the reflecting phase shifts of RIS 1 and RIS 2 as well as the 3D flight trajectory of the target UAV. We decompose the non-convex design problem into three subproblems and develop an iterative algorithm to find its approximated optimal solution by using the block coordinate descent technique. In each iteration, we utilize the successive con vex approximation, semidefinite relaxation, and phase alignment methods to handle these subproblems. Numerical simulation results are provided to validate the effectiveness and tremendous potential of dual-UAV-mounted RISs in the maritime sensing security.

In this paper, we study a cooperative game in the cooperative communication network, where each relay makes decisions autonomously and aims to achieve the same optimization objective of maximizing energy efficiency. We consider the non-ideal situation where instantaneous channel state information (CSI) is difficult to obtain and only partially observable outdated CSI is available. To solve this game problem, we define a delayed reward-based state-action value function and propose a multi-agent deep Q network learning framework. Then, we prove analytically that utilities obtained by game-theoretic approaches with the instantaneous CSI serve as upper bounds for those of the proposed method. Simulation results reveal that our approach considerably outperforms its potential alternatives and is only about 5.2

Decentralized Federated Learning (DFL) enables edge devices to perform collaborative model training in a distributed, peer-to-peer fashion, demonstrating significant advantages. However, DFL deployment encounters a core challenge: existing approaches cannot simultaneously achieve perfect consensus and communication efficiency in dynamic network environments. To address these challenges, this paper introduces over-the-air computation (AirComp) to boost communication efficiency by leveraging the natural superposition property of analog signals in wireless multiple access channels and employing efficient consensus algorithms to facilitate agreement and improve convergence accuracy. Specifically, we design and implement a novel multiple-input multiple-output (MIMO) BASE-GRAPH based AirComp-DFL (BA-DFL) framework to investigate the MIMO multiple access channel problem in over-the-air DFL under dynamic topologies. We conduct convergence analysis, with results encompassing both dynamic and static topological scenarios, reflecting the influence of dynamic topologies parameters and communication errors on MIMO OA-DFL performance in device-to-device (D2D) networks. The result indicates that subgraph length in topologies and communication errors significantly impact learning performance. Based on this, we formulate a comprehensive joint optimization approach that integrates communication and learning parameters to enhance overall system performance by simultaneously optimizing transceiver beamformers and dynamic network topologies. Extensive numerical simulations demonstrate the characteristic behaviors of various network structures and verify the significant improvement in learning performance achieved by our proposed algorithms.

To address the trajectory tracking control problem of nonlinear low-altitude aircraft under data packet loss, a data-driven model-free trajectory tracking control protocol is proposed. To circumvent the unmodeled dynamics introduced by practical complex constraints, the input-output data mapping of the system is established using dynamic linearization techniques. Based on this mechanism, a model-free adaptive control protocol is developed via a projection algorithm. Considering that data loss follows a Bernoulli distribution, a fault-tolerant control strategy based on dynamic compensation is designed. Furthermore, the convergence of the tracking error and the boundedness of the system's input and output are rigorously proved using the contraction mapping principle. Simulation results validate the effectiveness of the proposed compensation mechanism in enhancing tracking accuracy.

Equipping each antenna with an independent radio frequency (RF) chain in extremely large-scale antenna arrays (ELAA) requires unaffordable hardware and power consumptions. Analog beamforming is a common technology to reduce the number of RF chains. However, existing near-field positioning algorithms cannot adapt to analog beamforming, resulting in the algorithm being unusable when the RF chains are limited. This paper addresses the near-field localization problem for a base station (BS) with an ELAA and limited RF chains. Following the array partitioning approach, we divide the subarrays into groups and use a linear function to approximate the angles of arrival (AoA) variation within each group. We propose a message-passing algorithm, called array-partitioning-based location estimation with AoAs block-wise linearization (APLE-ABL), to estimate the user position. The simulation results show the effectiveness of our algorithm.