Extreme imbalance in sample ratio between traffic classes is an important factor affecting malicious traffic detection in Industrial Internet of Things (IIoT). Sample generation is an effective means to address the extreme imbalance malicious traffic detection. Traditional sample generation methods based on linear interpolation have difficulty in generating diverse samples for scarce traffic; while generation methods based on Generative Adversarial Networks (GANs) have difficulty in controlling sample generation boundaries, leading to the problem of generating overlapping traffic samples. To solve this problem, we propose a novel data enhancement algorithm called HybridGF. The method consists of three modules: hybrid generation module, filtering module, and explosion module. The hybrid generation module generates diversity samples adapted to the number of traffic category samples. The filtering module controls the boundary of generated samples by combining cluster analysis and Euclidean distance to avoid overlapping of generated traffic samples. The explosion module regulates the distribution of sample density in different clusters. In our experiments, we use three classical datasets (NSL-KDD, IoT-23, and TON-IoT) to simulate extremely unbalanced malicious traffic detection. The results show that HybridGF has a higher detection rate in detecting extremely imbalanced malicious traffic compared to other data enhancement algorithms. In the NSL-KDD dataset, the HybridGF model demonstrates a significant advantage in recall rate, with an overall improvement of 14.03% compared to the baseline model. Specifically, its recall rate for the DoS category increased by 4.6%.

Underwater sonar image target detection is essential for applications such as marine resource exploration, military defense, and underwater archaeology. However, current methods face difficulties in balancing detection accuracy and computational efficiency due to challenges like low resolution, severe noise, and ambiguous target boundaries. To address these issues, we propose a lightweight algorithm, Lightweight Sonar Detection Transformer (LS-DETR), which advances the trade-off between accuracy and efficiency through innovative module designs. LS-DETR consists of three core modules: Firstly, the RGLFN module is designed for efficient feature extraction by leveraging Ghost and RepConv mechanisms, reducing computational complexity while maintaining high-quality feature representations. Second, the CAMFPN module enhances multi-scale feature fusion and contextual modeling, addressing the challenge of robust feature extraction in high-noise sonar images. Third, the DPBAttention module introduces a dynamic position bias mechanism to model spatial relationships, significantly improving detection robustness in complex underwater environments. Extensive experiments conducted on the LASA dataset demonstrate the superior performance of LS-DETR. Compared to RT-DETR, a state-of-the-art real-time object detection model, LS-DETR reduces GFLOPs by 36% and parameter count by 28%, while maintaining comparable detection accuracy. Furthermore, LS-DETR outperforms the latest YOLO series model, YOLOv11, with mAP@50 increasing by 5.4% and mAP@50-95 increasing by 10.7%. These results establish LS-DETR as a new benchmark in underwater sonar target detection, achieving an optimal balance between computational efficiency and detection accuracy.

As a geological disaster widely distributed in the southern regions of China, rainfall-induced shallow landslides pose a significant threat to affected areas. Timely detection of landslides is crucial in the effective response to such disasters. However, landslide detection faces adverse impacts from various factors, such as insufficient sample data, complex model structures, and limitations in detection accuracy during the actual detection process. In this study, high-quality image samples were collected from multiple landslide disaster areas in southern China, and a rainfall-induced shallow landslide sample database was constructed in the region. Based on this, a lightweight attention-guided YOLO model (LA-YOLO) was proposed to improve the detection performance of YOLO model for rainfall-induced shallow landslides. First, CG block is introduced to enhance the C2f module, enriching the feature representation capability through multiscale feature fusion and reducing the model’s parameters and computational complexity. Second, the SimAM attention module is used to focus on the target regions, improving feature extraction effectiveness. Experimental results show that the model parameters of LA-YOLO were reduced by approximately 30%, with precision, recall, and mean average precision (mAP) on the landslide sample dataset increasing by 2.6%, 0.7%, and 2.2%, respectively. While ensuring model detection performance, the model structure was significantly optimized, achieving both lightweight and accuracy goals, confirming the model’s superiority in monitoring rainfall-induced shallow landslide disasters.

This paper addresses the critical challenge of achieving asymptotic stability in networked control systems (NCSs) under denial-of-service (DoS) attacks, focusing on maintaining security and stability within bandwidth-constrained environments. First, we construct a practical attack model using the NSL-KDD dataset to provide a realistic representation of DoS attack dynamics, capturing key attributes such as attack duration and frequency. Then, an iterative shrinkage-thresholding algorithm (ISTA) is introduced to supervise the adaptive sampled-data controller (ADSC), dynamically optimizing the sampling period to enhance control performance while minimizing communication overhead. To further mitigate the impact of DoS attacks, we propose a novel data compression mechanism that adapts to varying network conditions, ensuring efficient bandwidth utilization and preserving critical control data fidelity. In addition, the stability of the NCS is rigorously verified through Lyapunov-Krasovskii functions (LKFs), demonstrating robust system behavior even under adverse network conditions. Finally, the effectiveness and practicality of the proposed approach are validated through experimental studies on a 2-degree-of-freedom (2-DoF) helicopter system, confirming its capability to ensure stability, optimize communication efficiency, and mitigate the effects of DoS attacks in real-world scenarios.

Federated learning, as an emerging distributed machine learning paradigm, allows multiple parties to jointly train models without sharing raw data, thus solving the data silo problem and enhancing data privacy protection. As a powerful tool for processing graph-structured data, graph neural networks (GNNs) have shown great potential in many fields. Federated graph neural networks (FGNNs) combine the advantages of both, allowing data from different institutions to remain localized while utilizing GNNs to model complex graph-structured data, promoting multi-party collaborative training without directly exchanging data. However, the application of FGNNs in distributed environments faces many challenges, especially in protecting data privacy. This survey aims to comprehensively explore the privacy protection technology of FGNNs, covering the privacy protection mechanism of nodes, edges, and models, while also discussing the benefits and drawbacks of each approach. Finally, the future research directions of privacy protection technology in federated graph neural networks are discussed.

Cyber-attacks have become more frequent, targeted, and complex as the exponential growth in computer networks and the development of Internet of Things (IoT). Network intrusion detection system (NIDS) is an important and essential tool to protect network environments. However, the low performance of a NIDS against small malicious samples has seriously threatened the security of networks, thus directly leading to the loss of personal property and national interests. Given this, we propose an auto encoder-based hybrid detection model, abbreviated as AHDM, for the intrusion detection with small-sample problem. AHDM has a dual classifier framework. It trains first neural network based on the encoding features obtained from the autoencoder feature enhancement algorithm to detect small-sample malicious traffic. It trains second neural network using the original features to detect normal traffic and large-sample malicious traffic. The final detection result of malicious traffic is obtained by combining the detection results of the two neural networks. In experiments, we use three classic datasets (KDD CUP 99, CIC-IDS-2017, and IOT-23) and simulate the malicious traffic detection targeting extremely small-sample malicious traffic. The results show that AHDM has a higher detection rate for small-sample malicious traffic compared to the advanced detection models (DNN and ACID). In the IOT-23 dataset, the AHDM model shows an absolute advantage in detecting DDoS type of malicious traffic, with a detection rate of 0.71, which is much higher than the DNN (0.14) and ACID (0.14) models.

Unmanned aerial vehicle (UAV)-empowered edge computing has been widely investigated in obstacle-free scenarios, where a moving UAV is in charge of handling offloaded singleton tasks from mobile devices on the ground. However, little attention has been paid to the scenario, in which the UAV serves a complex area withmultiple obstaclesanddependent tasks. A dependent task can be formulated as a directed acyclic graph (DAG) that contains a number of subtasks; and each subtask can be executed by a corresponding service function (SF) deployed on the UAV. In this backdrop, the joint UAV trajectory planning, DAG task scheduling, and SF deployment is formulated as an optimization problem in this article. Afterwards, a deep reinforcement learning (DRL)-based algorithm is presented to tackle the NP-hard problem. The state space, action space, and the reward function of the agent, i.e., the UAV, are defined, respectively, under the DRL framework. To evaluate the effectiveness of the proposal, a series of experiments is conducted with different parameter settings. Results show that the DRL-based algorithm performs much better than three heuristic algorithms in success rate of trajectory planning, the number of executed tasks, and the average task response latency.

MCAM (CD146) is a cell surface adhesion molecule that has been reported to promote cancer development, progression and metastasis and is considered as a potential tumor biomarker and therapeutic target. However, inconsistent reports exist, and its clinical value is yet to be confirmed. Here we took advantage of several large genomic data collections (Genotype-Tissue Expression, The Cancer Genome Atlas and Cancer Cell Line Encyclopedia) and comprehensively analyzed MCAM expression in thousands of normal and cancer samples and cell lines along with their clinical phenotypes and drug response information. Our results show that MCAM is very highly expressed in large vessel tissues while majority of tissues have low or minimal expression. Its expression is dramatically increased in a few tumors but significantly decreased in most other tumors relative to their pairing normal tissues. Increased MCAM expression is associated with a higher tumor stage and worse patient survival for some less common tumors but not for major ones. Higher MCAM expression in primary tumors may be complicated by tumor-associated or normal stromal blood vessels yet its significance may differ from the one from cancer cells. MCAM expression is weakly associated with the response to a few small molecular drugs and the association with targeted anti-BRAF agents suggests its involvement in that pathway which warrants further investigation.

By employing time-varying proximal functions, adaptive subgradient methods (ADAGRAD) have improved the regret bound and been widely used in online learning and optimization. However, ADAGRAD with full matrix proximal functions (ADA-FULL) cannot deal with large-scale problems due to the impractical time and space complexities, though it has better performance when gradients are correlated. In this paper, we propose ADA-FD, an efficient variant of ADA-FULL based on a deterministic matrix sketching technique called frequent directions. Following ADA-FULL, we incorporate our ADA-FD into both primal-dual subgradient method and composite mirror descent method to develop two efficient methods. By maintaining and manipulating low-rank matrices, at each iteration, the space complexity is reduced from $O(d^2)$ to $O(\tau d)$ and the time complexity is reduced from $O(d^3)$ to $O(\tau^2d)$, where $d$ is the dimensionality of the data and $\tau \ll d$ is the sketching size. Theoretical analysis reveals that the regret of our methods is close to that of ADA-FULL as long as the outer product matrix of gradients is approximately low-rank. Experimental results show that our ADA-FD is comparable to ADA-FULL and outperforms other state-of-the-art algorithms in online convex optimization as well as in training convolutional neural networks (CNN).

In this paper, we try to draw attention to a particular type of force that has been somehow overlooked in the control of mechanical manipulators: gyroscopic force. At first, we broaden our understanding of gyroscopic effect intrinsic in mechanical systems by reformulating the Euler-Lagrange (EL) equation in terms of a gyroscopic part and the non-gyroscopic ones (or Rayleigh type). Then, we propose the use of artificial gyroscopic force to assist existing control laws. As a force that does not do any work, the gyroscopic force can be effectively used in combination with many existing control laws for mechanical systems without affecting their salient features such as the stability or the convergence to a target point. As a specific example, we look into the potential energy shaping parameterized by taskspace variables (also known as the Jacobian transpose method) as a base control law, and combine it with an artificial gyroscopic forcing term. We propose one specific way to design the gyroscopic force as a term quadratic in velocity that operates in reference to a desired velocity field shape (also parameterized in taskspace). The resulting control strategy enables us to shape the intermediate path profiles and thus realizes a taskspace control law that is free from any inverse transformations. We demonstrate the effectiveness of this method using simulation results with a three-link planar manipulator.

This paper proposes a method to recognize stroke orders of on-line handwritten Chinese characters based on analyzing both spatial and temporal information. A novel control-point-based analysis method is presented for spatial information analysis to match strokes with various shapes and styles. Its computation complexity is much lower than image correlation method and is suitable for applications on mobile devices. For temporal information analysis, Hidden Markov Model is adopted and a proposed rectification method is integrated to find the optimal pair-wise matching result of stroke sequences. Experimental results proved the effectiveness of the proposed method. The verification rate is 99.6% on the test set.