To address the challenges of information leakage, low energy efficiency, and the Doppler effect in mobile Internet of Vehicles (IoV), this paper proposes an enhanced IoV cooperation framework, where privacy information is forwarded by the untrusted relay assisted by unmanned aerial vehicle (UAV) and reconfigurable intelligent surface (RIS), which can improve security and energy efficiency. To meet the requirements of green communication, we formulate a secrecy energy efficiency maximization problem by jointly optimizing the transmit power allocation, the relay’s amplification factor, the two-hop RIS phase shift matrices, and the UAV trajectory. Given the non-convex nature of this problem, we introduce an iterative algorithm based on the convex-concave procedure and Dinkelbach’s method to optimize the transmit power and amplification factor. Additionally, we conceive the majorization-minimization (MM) algorithm to optimize the two-hop RIS phase shift matrices, and a designed firefly algorithm-deep deterministic policy gradient (FA-DDPG) algorithm is proposed to obtain the UAV trajectory. Simulation results demonstrate the effectiveness of the proposed scheme in enhancing secrecy energy efficiency. Specifically, compared to the DDPG-only and FA-based schemes, the proposed scheme achieves an improvement of 33.3% and 64.2%, respectively, in secrecy energy efficiency.

Reconfigurable intelligent surface (RIS)-enabled integrated sensing and communication (ISAC) is emerging as a key 6G technology for improving spectral efficiency and enabling high-resolution sensing. However,sensing targets within the communication coverage may act as potential eavesdroppers and intercept confidential data. To address this challenge, this paper proposes a sensing-assisted secure beamforming framework to enhance physical-layer security (PLS). First, we design a closed-loop architecture that sequentially performs RIS cascaded CSI estimation, target direction-of-arrival (DoA) estimation, and a Cramér–Rao bound (CRB)-based sensing accuracy evaluation. We then introduce an angular-domain information leakage (ADIL) metric to characterize leakage within the target’s angular uncertainty region. Building on this metric, we formulate a weighted-sum utility to jointly optimize the communication rate and sensing CRB under an ADIL-suppression constraint. To solve the resulting non-convex problem, we develop a penalty dual decomposition (PDD)-augmented alternating optimization (AO) algorithm that iteratively updates the BS beamforming, RIS phase shifts, and sensing time allocation. Convergence and complexity analyses further demonstrate that PDD accelerates AO convergence and mitigates zig-zag updates caused by coupled variables. Simulation results verify that the proposed sensing-assisted secure beamforming scheme effectively suppresses ADIL at eavesdropper angles and enhances PLS. Moreover, the PDD-augmented AO achieves up to a 21.3% improvement in communication rate and a 5.2% reduction in CRB compared with conventional schemes.

This paper investigates a novel RIS-assisted downlink mobile Internet of Vehicles (IoV) network to improve the service quality of long-distance communications. The proposed network leverages unmanned aerial vehicle (UAV) and reconfigurable intelligent surface (RIS), which can increase the signal path, thereby recovering the interrupted links and enhancing the signal strength. Additionally, the Doppler shift is considered in channel modeling to characterize the channel properties of dynamic environments. Aiming at the high-speed demand of 6 G networks, this paper introduces non-orthogonal multiple access (NOMA) technology to serve multiple vehicles and formulates a sum rates (SR) maximization problem. Then, an iterative framework is proposed for joint optimization, adopting the designed adaptive firefly algorithm (FA), semidefinite relaxation (SDR) technology, and deep deterministic policy gradient (DDPG) algorithm. Simulation results demonstrate that the proposed scheme can significantly enhance SR performance compared with space division multiple access and orthogonal multiple access schemes.

In this paper, we propose a novel physical-layer key generation (PKG) scheme for orthogonal frequency division multiplexing-orbital angular momentum (OFDM-OAM) systems to significantly enhance the confidentiality capacity (CC). In the proposed scheme, we first establish the OAM channel model under uniform circular array (UCA) misalignment in the line-of-sight (LoS) channel. Facing the risk of information leakage during key negotiation, we couple key generation with the OFDM communication process. Then, we analyze the CC of the OFDM-OAM system and derive its closed-form expression. Simulation results illustrate that the proposed OFDM-OAM PKG scheme has achieved high CC compared with existing works. In addition, as the offset angle of the eavesdropper’s UCA increases, the CC increases, and the bit error rate (BER) of the eavesdropper tends to be 0.5.

Mobile edge computing (MEC) utilizes reinforcement learning (RL) algorithms to optimize task scheduling, reducing latency and improving efficiency. Nevertheless, conventional RL approaches are constrained to single-action spaces and fail to adapt to dynamic user demands in vehicular edge computing networks (VECNs). To address these limitations, we develop a hybrid decision-making edge-assisted vehicular computation offloading (EVCO) model, which enables adaptive latency-energy balance, thereby enhancing quality of service (QoS). Specifically, we formulate a multi-objective optimization problem to minimize vehicular user costs by jointly optimizing latency and energy consumption. Furthermore, we propose the hybrid attention-enhanced deep deterministic policy gradient (HAE-DDPG) algorithm, which efficiently handles hybrid action spaces through a dual-layer network while dynamically adjusting optimization objectives via an attention mechanism. Field experiments demonstrate that HAE-DDPG enhances vehicular user cost optimization by 5-12% compared to baselines, while maintaining adaptability to dynamic environments.

In this paper, we investigate the secrecy performance of the UAV short-packet communication system assisted reconfigurable intelligent surface (RIS). In this system, based on the phase shift differences of the RIS, the Gamma and exponential distributions are used to match the received signal-to-noise ratio (SNR) at the link terminals. Closed-form expressions for both the probability density function (PDF) and the cumulative distribution function (CDF) are derived. Based on the above PDF and CDF, we derive closed-form expressions for the average achievable rate (ASR) and the average secure block-error rate (SBLER) to evaluate the system’s security and reliability performance. In addition, a novel analytical model is proposed, which can simplify the calculation of the secrecy outage probability (SOP). Furthermore, to explore the performance boundaries, we also derive closed-form expressions for the asymptotic SOP and the asymptotic probability of positive secrecy capacity (PPSC) in high-SNR regions. The accuracy of the derived expressions is validated through simulations and numerical results, which also demonstrate the effectiveness of the proposed SOP analysis framework. In the simulation, we investigate the impact of key system parameters, such as the number of RIS elements, finite block-length channels, UAV altitude, Rician factor, as well as reliability and confidentiality constraints, on the overall system performance. The simulation results show that the proposed system provides significant performance advantages in ensuring secure and reliable transmission under various operating conditions.

Class-incremental semantic segmentation focuses on updating the segmentation model with only new-class samples. Catastrophic forgetting and background shift are the two prevalent challenges. We identify two additional issues in remote sensing data that worsen these problems: significant class distribution variability and error accumulation-induced model degradation. To solve these three problems, we propose a new Self-Training and Curriculum Learning Guided Dynamic Refined Network (STCL-DRNet). First, we introduce a self-training auxiliary branch to complement the frozen last-step model, integrating cross-step knowledge to mitigate rapid forgetting. Then, a gradient-oriented Dynamic Refined Loss is proposed to assess under-learned classes and mitigate class imbalance. Furthermore, class-balanced curriculum learning is embedded to alleviate performance degradation throughout incremental training. Extensive experiments on benchmark datasets, including DeepGlobe, iSAID, ISPRS Potsdam, and Vaihingen, demonstrate that the proposed STCL-DRNet achieves state-of-the-art (SOTA) performance. In the 1-1s setting of the DeepGlobe dataset, STCL-DRNet exceeds previous SOTA methods by 11.6% in mIoU. For the iSAID 10-1s setting, it outperforms the previous SOTA by 12.76% in mIoU. As for ISPRS Potsdam and Vaihingen, our STCL-DRNet surpasses the SOTA by 5%-8% in all settings. Visualization and analysis further validate its interpretability. Our code is available at https://github.com/cv516Buaa/STCL-DRNet.

Cooperative positioning (CP) is able to improve the vehicular positioning performance by introducing the data of multiple vehicles into the position estimation. However, CP methods are vulnerable to measurement outliers in dense urban areas. The existing outlier-resistant CP methods are easy to trap in local optimum and may wrongly reject the outliers when the ratio of outliers to inliers is relatively high. To deal with this problem, a factor graph optimization (FGO) based CP method using Graduated Non-Convexity (GNC) Welsch cost is proposed in this paper. The state-of-art FGO algorithm is used to integrate multi-node and multi-epoch measurements including the Global Navigation Satellite Systems (GNSS) pseudoranges, inter-epoch baselines estimated by GNSS time-differenced carrier phase (TDCP), and inter-vehicle ranging measurements in a centralized framework. The least-square cost in traditional FGO is replaced with the GNC-based Welsch cost so as to enhance the robustness of the proposed method to any kind of outliers in our CP system. The use of GNC can reduce the risk of local optimum by gradually increasing the non-convexity of the Welsch cost. The proposed method can de-weight the outliers correctly even if a large number of outliers exist. The experimental results show the superiority of the proposed method over the existing CP methods in resisting multiple outliers.

Achieving higher accuracy in positioning at lower cost in urban environments is a fundamental requirement in modern intelligent transportation, autonomous driving, etc. The integration of low-cost inertial navigation system (INS) and global navigation satellite system (GNSS) is a promising choice to meet this objective, however, it still requires a suitable fault detection and correction method. Inspired by the structure of multi-channel parallelism in vector tracking, this paper introduces a multi-channel parallel processing (MCPP) method for identifying and eliminating GNSS faulty signals in the measurement domain in tightly integrated structure, aided by the principle of multiple model interaction. At each positioning epoch in each channel, the MCPP method individually compares the difference between each satellite’s measurement and its INS estimation version. These differences are normalized by a multiple model interaction process to derive the probability that each satellite is error free. These probabilities, linked to each satellite, can be used as weights for the optimal output or as indicators to identify false signals. The road test results in the Zhongguancun area of Beijing demonstrate that the method’s performance is comparable to that of state-of-the-art methods and avoids the unwanted outcome of selecting a fixed error elimination threshold that leads the system to diverge in a distinct environment or positioning mode. The results were valid in both tightly coupled (TC) and weighted least square (WLS) navigation modes, suggesting that the method can be utilized for single vehicle as well as vehicle-to-vehicle or vehicle-to-road cooperative positioning.

Unmanned aerial vehicle (UAVs) have the advantages of high flexibility and ease of deployment, making it possible to provide mobile edge computing services as an aerial server for remote or hot spot areas, e.g., computation offloading. However, there are bottlenecks in guaranteeing the reliability of computing resource allocation and incentivizing their participation in edge services. In view of this, we study the computation offloading and resource pricing joint optimization problem in the UAV-enabled vehicular edge computing network. In this article, we first formulate the interaction between vehicles and one UAV as a Stackelberg game, which maximizes the profits of the UAV and the utilities of vehicles considering delay, energy consumption, and urgency. Then, we analyze the existence and uniqueness of Stackelberg equilibrium (SE) under uniform and discriminatory pricing schemes applying backward induction. Finally, we implement such SE in both complete interaction information and incomplete interaction information scenarios. Specifically, one Stackelberg game-based dynamic iterative decision algorithm (SDID) and one reinforcement learning (RL)-based joint optimization offloading and pricing algorithm (RLOP) are proposed to intelligently obtain offloading and pricing strategies, respectively. Simulation results show that our proposed SDID and RLOP achieve significant improvements in the utility, compared to other baseline algorithms.

Incorporating an active reconfigurable intelligent surface on an autonomous aerial vehicles (AAVs), denoted as an aerial reconfigurable intelligent surface (ARIS), introduces a novel dimension for secure transmissions. Given the constraint of limited battery capacity in AAVs, energy management emerges as a key challenge within AAV networks. In response, we propose a secure energy efficiency (SEE) transmission scheme for ARIS networks, where active ARIS is strategically deployed to enhance information security. In addition, a SEE optimal problem is formulated by considering the imperfect wiretap channel state information to optimize the active beamforming vector and the ARIS position. For this non-convex problem, we first reformulate the fractional SEE objective into an equivalent form and subsequently decompose it into two distinct subproblems: optimizing the AAV’s position and designing the active beamforming. For the AAV’s position optimization, we propose a sophisticated deep deterministic policy gradient algorithm that enables the AAV to autonomously determine the optimal ARIS position through a self-learning strategy. Regarding beamforming design, we transform this aspect into a quadratic constrained quadratic programming problem and design an alternating direction multiplier method to optimize the reflection coefficient. Subsequently, an alternating optimization algorithm is proposed to synergistically solve these subproblems. Empirical simulations validate our proposed scheme, indicating an improvement in SEE of up to 47.2%. This significant improvement underscores the efficacy of the proposed ARIS-assisted secure transmission scheme in enhancing both security and energy efficiency in AAV networks.

In the context of rapid urban informatization, numerous vehicular devices have undertaken the responsibilities of local storage and data processing, with Federated Learning (FL) assuming a pivotal role within Vehicular Edge Computing Networks (VECNs). However, disparities in data quality and resources among vehicles may pose challenges to the efficiency of FL. To this end, we investigate the client selection and resource allocation issues specific to Unmanned Aerial Vehicle (UAV)-assisted vehicles within the domain of FL. Firstly, we construct a dynamic interactive reputation model where UAVs evaluate and select client vehicles based on factors like performance and capability, effectively filtering out high-quality data sources and enhancing the system’s ability to resist malicious node attacks. Secondly, we formulate a joint optimization problem to design a scheduling strategy that efficiently manages computational resources and communication capabilities, thus controlling latency and reducing energy consumption resulting from local model training. Additionally, we propose an asynchronous parallel Deep Deterministic Policy Gradient (APDDPG) algorithm with shared experience replay, aimed at enhancing the stability of global model convergence. Simulation results reveal that our proposed model and algorithm can more effectively resist attacks from malicious nodes and more fully utilize resources compared to other approaches, ultimately achieving efficient FL.

Class-incremental semantic segmentation aims to update the segmentation model with training samples containing only novel categories. Within this domain, catastrophic forgetting is a common challenge. In remote sensing scenes, images always have large discrepancies caused by a large variety of geographical objects. Therefore, besides catastrophic forgetting, there exists two additional primary challenges persist. The first one is a huge imbalance in image categories while the second one is error accumulation during multiple incremental training steps. To solve these three problems, we propose a new Self-Training and Curriculum Learning Guided Dynamic Refined Network (STCL-DRNet). Specifically, we first design a self-training-based branch to ease the tendency of catastrophic forgetting. We then design a dynamic refined loss to mitigate the uneven category distribution. Through further embedding class-balanced curriculum learning, we can alleviate the performance drop from noisy accumulation. Extensive experiments on benchmark datasets, DeepGLobe and iSAID, prove that the proposed STCL-DRNet achieves new SOTA performance. Visualization and analysis further substantiate the interpretability.

This paper proposes an active aerial reconfigurable intelligent surface (ARIS) assisted secure communication framework by integrating sensing and positioning against a mobile eavesdropper. In the proposed scheme, the base station (BS) beamforms the private information to the legitimate user and jams the eavesdropper with artificial noise (AN), while reconfiguring the phases and amplitudes of the passive signal by the active ARIS for promoting secure communications. To acquire the channel state information of the time-vary wiretap channel, the BS tracks the position of the eavesdropper by exploiting the reflected AN. Based on the tracked position of the eavesdropper in the previous time slot, we propose a secure communication scheme that aims to maximize the secrecy rate in the current time slot. This scheme is assisted by the ARIS through jointly optimizing the passive beamforming of the privacy information and AN, the reflection matrix of the ARIS, and the position of the ARIS. In the case of this non-convex quandary with highly coupled variables, we opt to disassemble it into three constituent subproblems and design an alternating optimization framework, where the optimal power beamforming at the BS is derived using a successive convex approximation method and semi-positive definite relaxation technique, the reconfigurable coefficient of the ARIS is optimized using the majorization-minimization algorithm, and the optimal position of the ARIS using the three-dimensional network is obtained by the deep deterministic policy gradient algorithm. Simulation results demonstrate the superior performance of the proposed scheme in the context of the secrecy rate when compared with benchmark schemes. By adopting the active beamforming and positioning technique, the secrecy rate can be increased by 38.3% and 10.8%, respectively.

Remote sensing scene zero-shot object detection aims to detect and recognize both seen and unseen catrgories of landscape elements with the guidance of the word embeddings. In this task, two primary challenges are identified. Firstly, there exists considerable variability within categories of landscape elements, causing a misalignment between visual features and word embeddings, particularly noticeable for unseen categories. Secondly, existing detection models struggle to provide accurate localization predictions, greatly impacting overall performance. To address these two issues, we propose WEA-DINO (Word Embedding Alignment-DINO). Based on the original DINO structure, our WEA-DINO-Head is specifically designed to align the hidden features of “matching queries” with word embedding features, effectively addressing the misalignment issue between visual features and word embeddings. Furthermore, aligning the hidden features of “denoising queries” with word embedding features enables the translation of localization capabilities from known categories to previously unseen ones. Through extensive experimentation on the DIOR benchmark dataset, our method demonstrates state-of-the-art performance. The code is available at https://github.com/cv516Buaa/WEA-DINO.

The advancement of deep neural network has propelled the widespread application of remote sensing target detection. However, compared to natural scenes, remote sensing targets possess inherent characteristics such as weak features and small scale, leading to a significant performance gap in traditional detection methods. To address these challenges, we undertake a systematic analysis of existing approaches, focusing on two key aspects: inadequate extraction of discriminative features and inappropriate regression measurement metrics. To tackle the first issue, an instance-level neighboring enhancement network (INEN) is proposed, enhancing the network’s feature extraction capability through inter-object feature aggregation. To address the second issue, a novel metric, smooth Wasserstein loss (SWL), is devised. Building upon these principles, a new tiny object detection (TOD) network for remote sensing images is developed. Extensive experiments on AI-TOD v1/v2 and DOTA v2 remote sensing tiny target detection datasets demonstrate that our approach achieves state-of-the-art (SOTA) performance. Codes are available athttps://github.com/sevenwgb/SWIN-TOD.

The development of vehicle-to-vehicle (V2V) communication facilitates the study of cooperative positioning (CP) techniques for vehicular applications. The CP methods can improve the positioning availability and accuracy by inter-vehicle ranging and data exchange between vehicles. However, the inter-vehicle ranging can be easily interrupted due to many factors such as obstacles in-between two cars. Without inter-vehicle ranging, the other cooperative data such as vehicle positions will be wasted, leading to performance degradation of range-based CP methods. To fully utilize the cooperative data and mitigate the impact of inter-vehicle ranging loss, a novel cooperative positioning method aided by plane constraints is proposed in this paper. The positioning results received from cooperative vehicles are used to construct the road plane for each vehicle. The plane parameters are then introduced into CP scheme to impose constraints on positioning solutions. The state-of-art factor graph optimization (FGO) algorithm is employed to integrate the plane constraints with raw data of Global Navigation Satellite Systems (GNSS) as well as inter-vehicle ranging measurements. The proposed CP method has the ability to resist the interruptions of inter-vehicle ranging since the plane constraints are computed by just using position-related data. A vehicle can still benefit from the position data of cooperative vehicles even if the inter-vehicle ranging is unavailable. The experimental results indicate the superiority of the proposed CP method in positioning performance over the existing methods, especially when the inter-ranging interruptions occur.

The deployment of satellite mobile edge computing (SMEC) incorporating artificial intelligence (AI) in low Earth orbit (LEO) constitutes satellite edge intelligence (SEI), which is promising to achieve autonomous processing of space missions on board driven by massive data. However, individual satellites with constrained resources and insufficient samples learn inefficiently, while the spatio-temporal constraints of large-scale LEO networks make collaborative training difficult. In this paper, a leader federated learning (FL) architecture for distributed SEI (SELFL) is proposed. By evaluating the connectivity and load of the dynamic constellation, the global and local parameters of the shared AI model are transmitted and updated continuously between the elected leader and other follower satellites based on the established inter-satellite link, which realizes efficient self-evolution of SELFL independent of the ground. Also we introduce a deep reinforcement learning-based resource allocation strategy for SELFL, which leverages the distributed proximal policy optimization (DPPO) to optimize the computing capability and transmit power of satellites for accelerating FL and reducing energy consumption. This method not only updates stably utilizing adaptive learning steps, but also improves sample efficiency with multiple parallel workers. The simulation results demonstrate the proposed SELFL optimization scheme effectively reduces the total energy consumption and training time by ensuring the AI model accuracy, and outperforms the benchmark algorithms.

Vehicular edge computing has emerged as a promising paradigm by offloading computation-intensive latency-sensitive tasks to mobile-edge computing (MEC) servers. However, it is difficult to provide users with excellent Quality-of-Service (QoS) by relying only on these server resources. Therefore, in this article, we propose to formulate the computation offloading policy based on deep reinforcement learning (DRL) in a vehicle-assisted vehicular edge computing network (VAEN) where idle resources of vehicles are deemed as edge resources. Specifically, each task is represented by a directed acyclic graph (DAG) and offloaded to edge nodes according to our proposed subtask scheduling priority algorithm. Further, we formalize the computation offloading problem under the constraints of candidate service vehicle models, which aims to minimize the long-term system cost, including delay and energy consumption. To this end, we propose a distributed computation offloading algorithm based on multiagent DRL (DCOM), where an improved actor–critic network (IACN) is devised to extract features, and a joint mechanism of prioritized experience replay and adaptive$n$-step learning (JMPA) is proposed to enhance learning efficiency. The numerical simulations demonstrate that, in VAEN scenario, DCOM achieves significant decrements in the latency and energy consumption compared with other advanced benchmark algorithms.

With the development of 5G and the Internet, the explosion of new mobile applications has led to an increasing number of computation-intensive and latency-sensitive tasks, which poses a severe challenge to resource-limited vehicles. Mobile Edge Computing (MEC) is recognized as an encouraging paradigm for providing offloading services from vehicles to the edge of wireless access networks. Computation offloading is the key technology in MEC that determines which tasks should be offloaded, for achieving the minimal time and energy consumption. However, in highway scenarios, statically deployed lightweight edge servers cannot meet the high resource requirements of vehicles with dependencies among subtasks. To alleviate this issue, this paper explores the utility of vehicles with idle resources as vehicular MEC servers. Considering the problem of dependency-aware task offloading, we utilize the directed acyclic graph (DAG) to analyze the task topology. Furthermore, due to privacy preservation, a distributed deep reinforcement learning-based algorithm with an optimized structure for offloading strategy is proposed. Convolutional neural networks and transformers are employed in the structure to extract rich state information efficiently. Experimental results reveal that, the proposed scheme achieves superior performance compared to the benchmark algorithms.

This paper presents a joint estimation and decoding algorithm for decoding over different impulsive noise in practical use, as the state of most impulsive noise changes rapidly in the wireless networks and cause performance degradation of low-density parity-check (LDPC) code decoding. The proposed algorithm fully utilizes the posteriori probability information of the decoding algorithm to strengthen estimation instead of regarding as the probability of input as equal probability. First, we derive the simplified objective of estimation expression. Then we designed the whole process of algorithm, which combines the estimator and decoder in an iterative manner. The simulation results show that the decoding performance of the proposed decoder can reach nearly known real noise parameters decoder within 0.1dB. Comparing with previous decoders, the proposed decoder has lower bit error rate (BER).