The accelerating global transition to electric vehicles (EVs) highlights the critical need for robust, accessible, and efficient charging infrastructure. Growing EV adoption leaves conventional charging networks grappling with fragmentation, user inconvenience, and suboptimal resource utilisation. This paper proposes a decentralised, blockchain-based framework for peer-to-peer (P2P) EV charging that leverages virtual zones to enhance scalability and efficiency. The framework enables a community-driven marketplace where users, charging point (CP) owners, and prosumers interact directly via smart contracts, eliminating the need for central intermediaries. A hierarchical matching service first performs rapid, local intra-zone Gale-Shapley matching to minimise wait and idle times, followed by inter-zone coordination to pair remaining unmatched entities across the network. To ensure real-world resilience, the framework integrates dynamic heuristics and a smart contract-driven reputation engine that mathematically detects and isolates adversarial or malicious nodes at runtime. The blockchain guarantees transparency, real-time tamper-proof CP status updates, and secure location verification, while a tiered matching system incentivises prosumer participation to foster network growth. By integrating hierarchical matching with distributed ledger technology, our framework addresses trust, scalability, interoperability, and dynamic resource allocation in community-based EV charging. Empirical evaluations on synthetic city-scale grids (up to 5,000 EVs) and real-world datasets (NYSERDA, Dundee City Council, ACN-Data) show that the architecture achieves 100% match, sub-second latency, and significantly outperforms legacy baselines under adversarial stress.
This article investigates the issue of dynamic event-triggered (DET) predefined-time cooperative formation target enclosed control for multiple vehicle systems, which subjects to infinite actuator faults and unknown dead-zones. The barrier Lyapunov function (BLF) is constructed to confine angular errors within predefined bounds, such that collision avoidance and inherent field-of-view restrictions of bearing sensors are ensured throughout the target enclosed control design. Meanwhile, considering the frequent updates of actual control signals, a DET mechanism is designed, which can significantly improve the utilization of communication resource. Combining the bearing measurement information and adaptive technique, a novel event-based robust adaptive cooperative formation target enclosed fault-tolerant control (FTC) algorithm is further developed. With the help of practical predefined-time stability theory, the developed enclosed control scheme can ensure that the multiple vehicle system is predefined-time stable and all signals are bounded. Finally, the effectiveness and feasibility of the presented control algorithm can be verified by simulation results.
This paper investigates the application of sparse superposition (SS) codes in multiple-input multiple-output (MIMO) systems, focusing on the joint mitigation of both transmitter-side and receiver-side nonlinear distortions—specifically, clipping and quantization. In contrast to prior works that address these distortions separately, we model them in a unified manner within a cascaded framework. An enhanced, replica-optimal orthogonal approximate message passing (OAMP) algorithm is developed for this integrated model. Simulations show that our OAMP-based approach effectively manages these nonlinearities simultaneously, yielding significant performance gains over alternative methods.
Integrated sensing and communications (ISAC) is revolutionizing low-altitude wireless networks (LAWN), making them vital for next-generation communication systems. However, effectively suppressing interference while balancing critical sensing tasks in multi-user multiple-input multiple-output (MU-MIMO) orthogonal frequency division multiplexing (OFDM) architectures remains a formidable challenge. By leveraging the spatial degrees of freedom (DoFs) in multi-antenna systems, this study presents an innovative transceiver beamforming design for MU-MIMO OFDM ISAC systems that effectively suppresses interference and optimizes beampattern sidelobes, significantly enhancing system efficiency. Specifically, we formulate a joint optimization problem to maximize the communication weighted sum rate subject to constraints on transmit power, the sensing signal-to-clutter-plus-noise ratio (SCNR), the peak sidelobe level (PSL) of the transmit beampattern, and spatial nulling requirements. To tackle the resulting NP-hard non-convex problem, we first transform it into an equivalent convex weighted minimum mean-square error (WMMSE) problem and develop an efficient alternating optimization algorithm based on successive convex approximation (SCA). Extensive simulation results demonstrate the effectiveness of the proposed scheme in efficiently suppressing interference and optimizing the beampattern, thereby confirming the efficacy of the proposed algorithm and paving the way for more reliable and efficient LAWN.
This article investigates the bit error rate (BER) and outage probability (OP) of multi-relay horizontal and vertical cooperative underwater wireless optical communication (UWOC) systems using selective decode-and-forward relaying. We consider a layered setup, with horizontal UWOC operating in a single layer and vertical UWOC spanning multiple layers. Beer–Lambert's law models absorption and scattering, while turbulence is modelled using an exponential generalized gamma distribution. Analytical expressions for average BER and an upper bound on OP are derived, along with high-signal-to-noise ratio asymptotic analysis and diversity order characterization. Monte Carlo simulations validate the analysis. The optimal source power allocation factor to achieve minimum BER under different relay configurations is also determined. We further examine the effects of temperature gradients and bubble levels (BLs) on BER and OP. Results show that increased turbulence due to higher temperature gradients, BLs, or both degrades UWOC performance, while deploying more relays in highly turbulent regions improves performance. For a vertical UWOC system with a temperature gradient difference of 0.10$^\circ$ C/cm, four relays reduce BER by 96.64% and OP by 89.4% compared with two relays.
The safety and stability of autonomous vehicles heavily rely on the proper functioning of their components. Distributed driven vehicles are particularly susceptible to actuator faults due to their multiple actuators, which elevate the risk of performance degradation and potential loss of vehicle control. This paper proposes a novel approach that integrates a robust trajectory tracking controller tolerant to both the steering actuator and in-wheel motors with reinforcement learning. Using a Linear Parameter Varying control framework, the proposed controller accounts for time-varying vehicle speeds, nonlinearities arising from tire behaviour and actuator faults. The controller relies only on data accessible from standard on-board sensors found in commercial vehicles and employs a static output- feedback control approach, ensuring practical implementation. System stability under faulty conditions is analyzed using the Lyapunov method, guaranteeing robustness to disturbances via $\mathscr {H}_{\infty }$ criteria. The controller performance is enhanced by a Deep Deterministic Policy Gradient reinforcement learning agent, which redistributes the wheel torques. The proposed methodology is evaluated across various driving scenarios under different fault conditions. The results demonstrate its robustness in handling faults during high-speed manoeuvres, where it maintains vehicle stability, achieving better tracking performance than other fault-tolerant approaches. Furthermore, the reinforcement learning agent further reduces tracking errors, enhancing the accuracy and performance beyond what the proposed robust fault-tolerant controller can achieve alone.
In emergency communication scenarios where unmanned aerial vehicles (UAVs) are employed as aerial platforms, limited resources such as time and power are extremely valuable, and service targets often exhibit heterogeneous priorities. To address the resource allocation problem for energy-efficient UAV-enabled integrated sensing and communication (ISAC) systems with differentiated service requirements, this paper proposes a joint optimization algorithm incorporating service target priority and the cooperation among communication, sensing, and energy. A priority model that jointly considers spatial location, task urgency, and target status is introduced to quantify the service profit of each target. A joint optimization problem involving target scheduling, transmit power allocation, and UAV trajectory design is formulated to maximize the UAV energy efficiency. Since the formulated problem is non-convex and difficult to solve directly, it is decomposed into three tractable convex subproblems, which are alternately optimized within a block coordinate descent (BCD) framework combined with successive convex approximation (SCA) to obtain a stationary solution. Simulation results demonstrate that the proposed algorithm significantly outperforms benchmark schemes in terms of energy efficiency. Furthermore, the proposed scheme enables differentiated services according to the importance of the targets, allocating more resources to high-priority targets while meeting the basic requirements of low-priority targets to maximize the system energy efficiency.
Distributed shape assembly enables multi-robot systems (MRSs) to autonomously reconfigure their spatial configurations, thereby achieving collective capabilities beyond those of individual robots. Existing methods often impose rigid interaction-topology constraints, limiting their adaptability to unstructured environments involving challenging terrain, sensing disruptions, and substantial communication delays. In contrast, this paper proposes a topology-free shape assembly framework. An attention-driven neural network is developed to compensate for transient errors in the consensus estimates used for generating motion commands before full convergence, thereby mitigating the effects of communication latency and accelerating shape formation. A negotiation-free dynamic goal-position assignment method is proposed based on local interactions within the communication range, enabling distributed shape formation without a predefined communication topology. Both simulation and real-world experiments demonstrate the feasibility of the proposed method under high-latency communication conditions relevant to vehicular scenarios with delayed vehicle-to-everything (V2X) links; in particular, compared with a representative baseline method, the proposed method reduces the shape-formation time by approximately 34.48% under high communication delays.
This paper investigates the outage performance of satellite-terrestrial networks (STNs) assisted by multiple aerial reconfigurable intelligent surfaces (RISs), implemented by mounting RISs on uncrewed aerial vehicles (UAVs). This configuration, termed a multi-RIS-UAV system, is analyzed under the joint effect of transceiver hardware impairments (HIs), RIS phase errors, and co-channel interference (CCI). In the proposed system, the UAV-mounted RISs perform passive beamforming toward a terrestrial user randomly located within the coverage area, thereby providing flexible coverage and enhanced link reliability. We propose a maximum received signal-to-noise ratio (SNR) RIS-UAV selection (MRUS) scheme to enhance system performance. By leveraging the Gamma distribution to approximate the composite fading of the cascaded satellite-RIS-terrestrial channel, we derive the exact outage probability expression for the MRUS scheme. Furthermore, asymptotic outage probability expression is obtained in the high SNR regime, through which the corresponding diversity order is characterized. Simulation results validate the analytical derivations and demonstrate the impact of key system parameters on performance.
In highly dynamic unmanned aerial vehicle (UAV) networks, the disintegration of traditional network boundaries poses a severe internal threat, thus enabling malicious UAVs to disrupt routing performances. To deal with the issue, we propose a secure multi-path routing framework built upon the zero-trust architecture and software-defined networking, empowering the global resource management and policy implementation. To mitigate the damage of malicious behaviors, a multi-dimensional dynamic trust evaluation model is designed, integrating factors of the data interaction frequency, success rate, certificate validity, and location compliance. Furthermore, through continuous scoring for trust values and the dynamic authorization of UAVs, zero-trust principles are effectively integrated into the routing framework for abnormal and non-compliant node management. Moreover, we formulate the routing problem to jointly optimize the end-to-end delay and transmission loss rate, which is an nonlinear programming problem that is intractable to solve. To address this, we reformulate it as a Markov decision process and propose a link-state adaptive deep reinforcement learning algorithm tailored for highly dynamic UAV networks. Numerous simulations are conducted and results show that the proposed mechanism can effectively ensure lower delay.
In this paper, we investigate fingerprint-based localization in high-dynamic environments for cell-free massive MIMO-OFDM systems and propose CLALoc, a long-term and adaptive fingerprint localization scheme. We begin by constructing the fingerprint based on the angle-delay domain channel power matrix, which encapsulates critical localization-based channel characteristics. Subsequently, crowdsourced fingerprints are utilized through semi-supervised transfer learning to automatically infer reliable location labels, allowing the fingerprint database to grow without costly manual collection. As environmental variations emerge over time, a meta-learning based adaptation mechanism enables rapid model adjustment with only a few newly collected fingerprints. To ensure stability during continual updates, gradient episodic memory is employed to preserve representative historical knowledge and mitigates catastrophic forgetting. Simulation results based on dynamic channel datasets generated by DeepMIMO and Wireless InSite demonstrate that CLALoc significantly improves localization accuracy and robustness compared with existing fingerprint localization methods, particularly in long-term dynamic environments.
The rapid expansion of global industrial infrastructure and modular construction has led to an unprecedented demand for the logistics of oversized cargo. Electrically driven self-propelled modular transporters (e-SPMTs) indicate significant advantages in oversized cargo delivery for both national defense security and social infrastructure due to inherent flexibility and high mobility efficiency, especially for long-distance transportation under complex road conditions. However, coordinating multiple e-SPMTs introduces nontrivial challenges: Trajectory planning on confined roads must account for collision avoidance and coupled dynamics, while the redundancy and nonlinearity of steer-by-wire systems further complicate the design of effective control strategies. To this end, this paper proposes a comprehensive framework comprising a two-stage trajectory plan strategy and a hierarchical control scheme. Specifically, the kinematic and dynamic models of the coordinated system are established to capture the rigid physical coupling effects. Subsequently, the planner generates a curvature-continuous and kinematically feasible reference trajectory via integrating safety corridor concepts into the two-stage optimization strategy. To track these reference signals, a hierarchical control architecture is designed, featuring an upper-layer tracking controller and a lower-layer tire force distributor, to manage actuator redundancy and further ensure lateral stability. Finally, co-simulation and hardware-in-the-loop tests under complex transportation scenarios demonstrate the feasibility and effectiveness of the proposed framework and approaches.
The Signal of OPportunity (SoOP) from Low-Earth-Orbit (LEO) satellites, characterized by spectral and geometric diversity, high received power, and rapid orbital motion, is used to assist the Global Navigation Satellite Systems (GNSS) positioning in GNSS-obscured and GNSS-denied scenarios. We propose an LEO-aided GNSS functional model. Thereafter, the Cramer-Rao Lower Bound (CRLB) is derived to assess the feasibility of the LEO-aided GNSS positioning. To address the unknown stochastic characteristics of LEO signals, we propose an Adjusted Factor Graph Optimization (AFGO) algorithm that fuses GNSS and LEO observations to enhance positioning performance in obscured and denied GNSS scenarios. Moreover, the computational complexity is analyzed to highlight the limitations of the AFGO algorithm. Finally, the four joint GPS and LEO field experiments are conducted to evaluate the positioning performance of different algorithms across various application scenarios under the LEO-aided GNSS framework. The experimental results show that the positioning performance of the AFGO algorithm is better than that of the extended Kalman filter and adjusted moving horizon estimation in terms of convergence, precision, and accuracy under insufficient GNSS satellites. The AFGO algorithm employs LEO satellites to augment inadequate GNSS observations, thereby narrowing the search area in forest rescue scenarios.
This paper proposes an event-triggered finite-time model-free adaptive PID control scheme for high-speed train speed regulation under nonlinear and time-varying dynamics. First, an equivalent linear data model is established using input-output data, and a finite-time model-free adaptive PID controller is developed that guarantees tracking performance without requiring an explicit system model. Second, a dual-channel packet-loss compensation strategy is formulated to mitigate data dropouts in both the forward and feedback communication channels. Third, an event-triggered communication mechanism updates the control law only when an error-driven condition is satisfied, thereby reducing communication overhead. Finally, a rigorous analysis of tracking-error convergence is presented, and simulation studies validate the effectiveness and superiority of the proposed scheme.
With the rapid advancement of the low-altitude economy and low-altitude wireless networks (LAWNs), unmanned aerial vehicle-to-vehicle (U2V) communication has become a critical enabler for emerging applications such as urban air mobility, logistics, and intelligent transportation. Existing airto- ground channel studies predominantly assume static or quasistatic ground terminals, thereby neglecting the influence induced by ground-terminal mobility. This gap motivates dedicated U2V channel investigations under realistic dynamic conditions. This paper presents a comprehensive U2V channel measurement campaign conducted in a representative mixed-urban environment, where both the UAV and the ground vehicle are in motion. Based on the measurement data, the time-frequency non-stationary characteristics are first quantified, followed by analyses of smalland large-scale fading features, including envelope distributions and line-of-sight (LOS) probability. An environment-aware U2V path-loss model incorporating LOS probability is developed to capture severe propagation fluctuations. Power delay domain parameters and channel sparsity metrics are further examined across different scenarios. The results reveal that U2V channels in mixed-urban environments exhibit rapid fading transitions governed by dynamic scatterers such as vehicles and buildings, and that the proposed sparsity metrics effectively characterize the structural evolution of multipath components. These findings provide essential guidance for the design and optimization of future LAWNs and vehicular communication systems.
In scalable cell-free massive MIMO (CF-mMIMO) systems with hardware impairments (HWIs), efficient precoding design is crucial for maximizing energy efficiency (EE) under power and quality of service (QoS) constraints. However, conventional optimization-based approaches face challenges in scalability and complexity, while existing learning-based methods struggle to balance performance and generalization. To address these challenges, we develop hardware-aware (HA) precoding, along with tailored EE optimization and learning strategies. First, leveraging uplink-downlink duality, we design HA precoders and analyze their dependence on HWI factors. Considering that the formulated precoding optimization problem cannot be directly transformed into real-valued forms, we propose a complex projected gradient method with a closed-form solution for the projection step. Furthermore, we introduce a low-rank (LR) Transformer-based neural network that integrates supervised and unsupervised learning for implicit joint optimization of virtual uplink powers and downlink precoding vectors. Simulation results validate that the proposed methods achieve substantial EE gains, reduced computational complexity, and strong scalability.
Numerology ($\mu$) is the key design parameter of the physical layer in Fifth Generation New Radio (5G NR) networks and beyond, including Vehicle-to-everything networks (V2X). $\mu$ shapes communication link specifications, such as Spectral Efficiency (SE) and link rate. The subcarrier spacing $\Delta f$ and the guard interval $T_{g}$ mitigate the effects of mobility and multipath, respectively. Relying on the functionality of $\Delta f$ and $T_{g}$, this paper proposes a novel adaptive numerology technique for maximizing the SE in 5 G NR networks and beyond, including V2X networks (ANV2X). ANV2X is a channel-aware adaptive numerology technique that links channel parameters, including mean, maximum, root mean square (RMS) delay, and Doppler spread for numerology selection (allocation), represented by $\Delta f$ and $T_{g}$. ANV2X introduces exact bounds on $\Delta f$ and $T_{g}$, revealing a crucial result that $\Delta f$ and $T_{g}$ are highly coupled with the channel parameters. Extensive simulation results demonstrate that ANV2X outperforms state-of-the-art adaptive numerology systems and Fixed Numerology Systems (FNS) for SE optimization across various frequency ranges. For example, in Case 1, Sub-6GHz system, ANV2X enhances the SE by 7.005×, 6.5624×, 3.2986×, and 0.3603× compared to ANS, ASB, FNS systems with $\mu _{0}$, $\mu _{1}$, and $\mu _{2}$, respectively. ANV2X has the potential to be a new technique for 5 G NR networks and beyond, as it characterizes the commonality of links, including adaptive numerology and a flexible physical layer design.
Orthogonal time frequency space (OTFS) modulation offers strong resilience to Doppler effects but suffers from high system latency, limiting its use in low-latency communications. This paper proposes a channel estimation algorithm for low-latency OTFS systems with large delay spreads and fractional Doppler effects. In the delay-time (DT) domain, impulse pilots are placed at equidistant intervals along the first row of the DT grid to eliminate interference from aliased delays, and a threshold detection method estimates the delays. Doppler shifts and path gains are then estimated using the discrete Fourier transform (DFT). Simulation results show that the proposed algorithm achieves near-optimal bit error rate (BER) under sparse delay spreads.
The Space-Air-Ground Integrated Network (SAGIN) has emerged as a pivotal architecture for sixth-generation (6G) mobile communication systems, offering extensive multi-domain and multi-dimensional coverage. To support intelligent services within SAGIN, deep neural network (DNN) inference tasks are often distributed across multiple devices due to the increasing scale of DNN parameters and constrained communication and computation resources. In this paper, to enhance intelligent services in the Internet of Vehicles (IoV), a cognitive-radio (CR)-enhanced SAGIN framework is proposed, where unmanned aerial vehicles (UAVs) are responsible for data collection and cooperate with vehicles to complete complex DNN inference tasks. A joint optimization problem that coordinates device association, model splitting with early exiting, resource allocation, and UAV trajectory planning is proposed to improve the average inference efficiency over a finite multi-round mission, defined as the average of the per-round ratios between aggregate inference accuracy and aggregate inference latency. Since the problem is highly coupled and non-convex, we decompose it into several subproblems and solve them iteratively. A quadratic unconstrained binary optimization (QUBO)-based graph attention network (GAT) approach with independent branch selection is proposed to solve the joint device association and model splitting problem, where a differentiable loss function is derived through Hamiltonian relaxation. In addition, we obtain a convex approximation of the transmit-power subproblem, while computation-resource allocation is solved as a convex optimization problem. For multi-round operations, we design a multi-agent reinforcement learning (MARL)-based trajectory optimization algorithm under a centralized training and decentralized execution paradigm. Simulation results on widely used DNN models, including AlexNet and VGG16, demonstrate that the proposed approach can improve the efficiency of collaborative DNN inference.
This correspondence investigates the secrecy energy efficiency (SEE) issue with the assistance of a reconfigurable intelligent surface (RIS). More particularly, we consider the downlink communications of multiple-input-multiple-output wireless networks, where multiple active colluding and non-colluding eavesdroppers (Es) attempt to wiretap the legitimate transmissions. To evaluate the SEE performance of the system, we introduce two distinct schemes: the SEE maximization based colluding eavesdropping (SEEM-CE) scheme and the SEE maximization based non-colluding eavesdropping (SEEM-NCE) scheme. Then, the proposed optimization problem is decomposed into two sub-problems, namely the base station (BS) precoding vectors optimization sub-problem and the RIS's phase shifters adjustment sub-problem. To be specific, for the first sub-problem, the BS precoding vectors are derived through the Dinkelbach's method and successive convex approximation (SCA) strategy. For the second sub-problem, an SCA-based approach is developed to optimize the RIS phase shifts. Building on these solutions, a block coordinate descent algorithm is proposed to alternately optimize the BS precoding vectors and RIS phase shifts, thereby addressing the original problem. Finally, numerical simulations demonstrate that the proposed SEEM-NCE algorithm achieves superior performance compared to the SEEM-CE scheme and other benchmark methods.
