Magnetic particle imaging technology, a novel medical imaging technology, possesses rapid imaging, high penetration depth, and is free from ionizing radiation. However, the system point spread function causes imaging blurring, which can be further exacerbated by external environmental interferences. Although hardware improvements and system optimization can mitigate blurring, these approaches are often expensive and time-consuming, particularly for low-field imaging in large-scale systems. This article proposes a Fast Context-aware Saliency-enhanced Deblurring Network, FCS-edNET, to solve the challenging issue by deblurring the reconstructed images. The network introduces the Multi-scale Global module to enhance the multi-scale feature perception ability. The Multi-scale Denoising Prior algorithm, which employs a low-frequency filter operator to restrict image noise and offers priors for each layer of subnetworks, is designed to improve the model robustness. Finally, proposing a Multi-level Joint loss optimizes model parameters to promote model convergence speed and space distribution simulation capability. Extensive experiments on multiple public and private datasets demonstrate that FCS-edNET outperforms the state-of-the-art methods in MPI image deblurring efficiently, suggesting its potential to support future research toward clinical imaging applications. The code is available at https://github.com/ydz1118/FCS-edNET.

High-altitude airships (HAS) and unmanned aerial vehicles (UAVs) equipped with Multiaccess Edge Computing (MEC) servers have emerged as promising aerial MEC nodes for providing task offloading (TO) services to intelligent mobile devices (IMDs) in post-disaster scenarios. HAS offers robust computing and energy resources, while UAVs provide flexible, low-altitude coverage for rapid deployment. However, direct task offloading from IMDs to HAS often leads to task failures due to high transmission delays. UAVs with limited onboard resources require to minimize resource waste. Additionally, IMDs in sparse areas face insufficient TO services due to unfair UAV coverage. This paper defines these challenges as a joint optimization problem involving TO, RA, and UAV coverage fairness. It proposes a cooperative aerial Multiaccess Edge Computing (AMEC) framework integrating HAS and UAVs to address the issue. Within this framework, a hybrid TO scheme is first developed to mitigate the high transmission delay between IMDs and HAS. Second, a Distance, Resource, Urgency-based Decision Mechanism (DRUDM) is designed to enhance the accuracy of UAVs in selecting target IMDs for TO services. Third, a Coverage Fairness Guarantee (CFG) strategy is proposed to optimize UAV flight trajectories, ensuring IMDs in sparse areas receive fair TO services. Finally, the joint optimization problem is modeled as a Multi-Agent Partially Observable Markov Decision Process (MA-POMDP), and a DRUDM–CFG algorithm is presented to efficiently solve this complex non convex optimization problem. Experimental results demonstrate that the proposed algorithm outperforms other compared algorithms in task completion rate and average delay, benefiting from the DRUDM mechanism. Meanwhile, the CFG strategy effectively improves TO service fairness for IMDs in sparse areas.

In the Internet of Autonomous Vehicles (IoAV), task offloading is crucial for managing tasks that require extensive computing power to guarantee vehicle safety under different weather scenarios. However, extreme weather events can lead to infrastructure damage and network disruptions, significantly increasing the computational demands of autonomous vehicles. These vehicles require additional computing resources to navigate complex road conditions and risks, all while facing a high degree of uncertainty, such as fluctuations in vehicle resource utilization and task workloads. To address these challenges, a new and lightweight task offloading decision framework, named 2FDP-BRL, has been first proposed in this paper. This framework not only considers the fast response time required for autonomous driving, but also considers the resource shortage and offloading uncertainty caused by extreme weather. Therefore, we introduce the dynamic pricing idea and the Interval Type-2 Fuzzy Inference System (IT2FIS) utilizing broad reinforcement learning to deal with various dynamic uncertainties in the IoAV under extreme weather. For the authenticity of experimental results, we utilize the VISSIM platform to collect experimental data and conduct simulations. Moreover, to accurately simulate extreme weather scenarios, we also account for the variability of infrastructure and road elements, including reduced transmission rates and decreased efficiency in executing tasks. Furthermore, to enhance the realism of the simulation, we incorporate historical weather data from NOAA for Shenyang in 2024 to model dynamic uncertainties under extreme weather conditions and conduct comparative experimental analyses focusing on task completion rates. Finally, the proposed framework was implemented on both a local setup and the Huawei Atlas 200I DK A2 device, illustrating its efficacy design.

Person re-identification (Re-ID) aims to match individuals across different cameras, a task complicated by variations in camera positions, resolutions, and lighting conditions. While supervised training improves Re-ID model accuracy, it requires significant annotation efforts. Unsupervised domain adaptation (UDA) methods address this by leveraging unlabeled target domain data but often fail to fully utilize multiple source domains and are constrained by computational resources. This article introduces a lightweight multi-source domain adaptation method for person Re-ID that combines meta-learning with pseudo-label-based UDA. By employing Self-paced Meta-Learning (SpML) and style enhancement techniques, the model learns domain-invariant knowledge from easy to difficult source domains, enhancing pseudo-label quality during adaptation. Our approach, based on an omni-scale feature extraction network using deep separable convolution, combines global and partial feature branches to capture richer pedestrian features. Experiments on public and real-world datasets demonstrate that our method achieves competitive performance with significantly fewer parameters and Floating Point Operations (FLOPs) compared to state-of-the-art models, proving its effectiveness and practicality.

In mechanical manufacturing and assembly, precise alignment of the spindle and impeller is essential for improving the efficiency and reliability of automated production lines. With the development of IoT technologies, high-resolution 3D scanning systems enable real-time acquisition of point cloud data, supporting accurate geometric modeling and intelligent alignment. However, traditional RANSAC-based cylindrical fitting methods often suffer from poor robustness and accuracy in complex, noisy environments. This paper proposes a Dual-Consensus RANSAC (DC-RANSAC) algorithm that enhances fitting reliability by fusing global consensus degree—which evaluates the overall agreement of point samples—and local spatial density consistency—which captures the structural coherence of inlier neighborhoods. This dual mechanism addresses the limitations of conventional inlier-count-based evaluations by suppressing pseudo-inlier effects and improving model integrity. Experimental results demonstrate that the proposed algorithm significantly outperforms conventional RANSAC in both fitting accuracy and noise robustness, effectively reducing alignment errors in practical spindle-impeller assemblies.

Edge-Level Graph Learning System (EGLS) exhibits diverse applicability in management of distributed vehicular networks, e.g., flow prediction, route planning, and accident forecasting. For the EGLS training, expensive hardware resource consumption, traffic data collection, and dedicated training procedures make the learning algorithms become valuable intellectual property (IP) for the EGLS owner (e.g., Uber and Lyft), and they cannot tolerate the infringement act of their models’ intellectual property. To enhance its IP protection, we present ActIPP, the first active IP protection methodology for EGLS, which incorporates a built-in access control function in the model to safeguard against unauthorized queries. Specifically, it is achieved via a creative edge backdoor mechanism, wherein the edge training samples are poisoned via user-specific access tokens to induce legal outputs from a well-trained EGLS model for authorized users. Moreover, related token regulating strategies were proposed to dynamically realize the addition and revocation of user tokens by model retraining to guarantee access control in EGLS. Additionally, a Graph Mutual Information-based adaptive token generation method is presented to augment the access control embedding. Based on experiments with various real-world datasets, ActIPP demonstrates high success rates of IP protection (accuracy drop < 4%) under various scenarios and efficiently prevents unauthorized access (unauthorized access accuracy < 6%).

In the Internet of Autonomous Vehicles (IoAV), task offloading is a method to address computationally intensive tasks and ensure the safe operation of vehicles. However, under extreme weather conditions, the number of these tasks significantly increases, posing higher risks and challenges. Therefore, to mitigate risks and ensure the safe operation of vehicles, it is crucial to make quick and effective decisions during the task offloading process. Currently, most methods in this domain utilize Deep Reinforcement Learning (DRL). However, the large number of parameters in deep networks results in the characteristics of long decision time and large consumption of computational resources. In order to solve this problem, this paper proposes a task offloading scheme named Dynamic Pricing Driven Double Broad Reinforcement Learning (DP-DBRL), which utilizes Double Broad Reinforcement Learning (DBRL) to reduce model memory consumption and decision time. Additionally, it considers the high-speed mobility and resource variability to devise a more efficient dynamic pricing scheme that minimizes the overall delay in task processing for vehicles. To validate the proposed scheme, we conduct simulations using the VISSIM platform, meanwhile, we simulate task offloading scenarios under extreme weather conditions by randomly reducing factors such as the transmission rate and task execution efficiency of infrastructure and vehicles on the road. Finally, we deployed the proposed scheme both locally and on the Huawei Atlas 500 device to demonstrate its effectiveness and lightweight nature.

Autonomous driving has brought about a growing interest in enhancing traffic efficiency and ensuring road safety. One of the fundamental functions of autonomous driving technology is lane-keeping, which has become a popular research topic in autonomous driving. However, current deep reinforcement learning (DRL)-based algorithms used for solving lane-keeping problems have limitations, such as low sample utilization and high time cost in complex scenarios. To address this, we propose a lane-keeping transfer Double Deep Q Network (LK-TDDQN) framework that leverages transfer learning (TL). Our framework enables autonomous vehicles to perform lane-keeping tasks in similar scenarios, transferring knowledge from a single-lane rural road scenario to a two-lane racing scenario. The effectiveness of the proposed LK-TDDQN was demonstrated in several simulation experiments in OpenAI Gym. These simulations demonstrate that our approach can enhance decision-making efficiency by 22% and reduce the time cost of autonomous vehicles by 11%, ensuring the safety of autonomous driving while alleviating the burden on drivers.

Recently, vehicle classification is becoming increasingly important with the development of automated driving technology. In particular, it can provide the basis and prerequisites for autonomous vehicles to make decisions in terms of improving driving safety. However, the current mainstream vehicle classification methods are deep learning algorithms based on Convolutional Neural Networks (CNN), which are mainly focused on the cloud, and these algorithms have complex models and large training parameters. In addition, for computationally intensive and urgent tasks, the poor computational power and low storage capabilities of edge nodes cannot support CNN-based vehicle classification algorithms for model updating. In this paper, we propose a lightweight vehicle classification method with mobile edge computing based on Broad Learning System (BLS). On the one hand, the vehicle can serve as a mobile edge computing node to provide computing and storage resources to ensure that classification tasks are performed locally and quickly, avoiding the bandwidth congestion caused by uploading to the cloud. On the other hand, we use broad learning method to perform incremental training on the data, which is more suitable for computing at the edge, because it can support incremental updates to the model on the vehicular edge nodes without retraining the whole model. Experiments are conducted on a Raspberry Pi system to simulate edge nodes, the results show with a similar performance, the training speed of our vehicle classification system can be increased by 10 times compared with the other CNN-based algorithms.

Vehicular fog computing (VFC) could provide fast task processing services for vehicles. To make vehicles/fog nodes willing to buy/sell resources, a double auction mechanism considering the interests of all parties is needed. However, few works study the auction issue in VFC. Different from the existing edge-related auction which only considers the price, some nonprice attributes (location, reputation, and computing power) are also important for providing fair resource allocation in VFC. In this article, we propose a multiattribute-based double auction mechanism in VFC, which considers both the price and nonprice attributes for constructing reasonable matching. To the best of our knowledge, this is the first work to consider multiattribute-based auction in VFC. Our auction mechanism could satisfy computational efficiency, individual rationality, budget balance, and truthfulness. To verify the proposed mechanism, we simulate VFC using VISSIM and extract the driving data. The experimental results show the effectiveness and efficiency of this mechanism.