The rapid expansion of electric vehicles (EVs) charging stations underscores the urgent need for robust anomaly detection systems capable of identifying potential malfunctions while preserving data privacy. Federated Learning (FL) has emerged as a promising solution, enabling collaborative model training without requiring raw data sharing. However, applying conventional FL approaches to charging station networks presents significant challenges, including non-independent and identically distributed (non-IID) data and gradient conflicts among clients. To address these challenges, we introduce FedPareto, a novel Pareto-optimal FL framework designed to manage gradient conflicts and counter malicious attacks in charging station anomaly detection. FedPareto features a gradient conflict-aware aggregation method, which adaptively adjusts client weights based on cosine similarity between gradients, and a gradient magnitude reshaping strategy to enhance model convergence. Theoretical analysis demonstrates that FedPareto achieves a convergence rate of O($\frac{1}{T}$) and attains Pareto-optimal solutions under standard smoothness and convexity assumptions. Extensive experiments on real-world charging station datasets validate FedPareto’s effectiveness. It outperforms state-of-the-art methods, exhibits better robustness against gradient-based attacks, and ensures equitable performance distribution across clients. These results highlight FedPareto’s potential as a reliable and scalable solution for anomaly detection in EV charging station networks.

federated graph learning (FGL) has emerged as a promising paradigm for privacy-preserving collaborative learning in Social Internet of Things (SIoT), where nodes form complex interconnected networks. Existing FGL approaches face significant challenges including model degradation in handling nonindependent and identically distributed (non-IID) data and maintaining model performance across heterogeneous nodes. This article proposes framework via layer-wised clustering (FedLWC), a novel layer-wised clustering framework inspired by evolutionary processes is proposed to enhance the effectiveness of FGL. FedLWC designs three key aspects: 1) a fisher information matrix-based layer selection mechanism that identifies and evaluates critical model layers, which can reduce parameter redundancy; 2) a layer intersection clustering algorithm that preserves common key layers while accommodating local features; and 3) an adaptive layer merge strategy that effectively combines global shared layers with clustered key layers. To make sure that the proposed approach is rigorous, we conduct theoretical convergence analysis for the proposed framework under non-IID conditions. Extensive experiments on multiple benchmark graph datasets demonstrate FedLWC’s performance, achieving an average accuracy improvement of 7.01% compared to state-of-the-art federated learning methods.

Federated learning (FL) faces significant challenges in scenarios with nonindependent and identically distributed (non-IID) data distributions across participating clients. Traditional aggregation-based approaches often struggle with the inherent misalignment between local and global optimization objectives, which leads to gradient divergence and suboptimal generalization performance. This article proposes a novel FL framework that replaces conventional aggregation with a biologically inspired model symbiosis approach called FedSym, which employs a dual-level symbiotic mechanism. Ectosymbiosis performs coarse-grained hierarchical parameter recombinations through random layer-wise model combination, while endosymbiosis enables fine-grained intralayer parameter fusion through weighted averaging, collectively steering model updates toward flatter loss landscapes. Our theoretical analysis demonstrates that FedSym’s convergence rate is$O({}{1}/{T})$under non-IID conditions, which matches the convergence properties of FedAvg. Extensive evaluations across multiple datasets and model architectures show that FedSym achieves substantial improvements over state-of-the-art FL methods, particularly in challenging scenarios with high data heterogeneity, and demonstrates robust performance across varying numbers of participating clients and federation scales.

Fault diagnosis is important for predictive maintenance in smart manufacturing, which involves intelligent human-machine interactions in order to make smart decisions for potential problems. Large language model (LLM) is promising in providing general artificial intelligence capabilities in this regard. However, LLM itself can not accurately analyze faults due to heterogeneous data from different Industrial Internet of Things (IIoT) devices in different processes during the complete production process. To accurately diagnose faults and facilitate human-machine interaction, this article proposes a brain-like cognition-driven model factory (BC-MF), using an LLM as a supervisor to adaptively generate personalized small-scale models according to the features of these heterogeneous data, where the vertical federated learning (VFL) idea is adopted. This BC-MF-based fault diagnosis approach includes a preliminary diagnosis phase and a precise diagnosis phase. The preliminary diagnosis is accomplished by prompting the LLM using a brain-like chain of thoughts (BLCoTs). A hypernetwork uses the preliminary diagnostic results and the feature maps trained by each node in the VFL to generate dedicated diagnostic small models and uses these models for final precise diagnostics. The LLM provides fault maintenance recommendations interactively according to the final diagnostic results. Comprehensive evaluations are conducted using four open IIoT datasets and one self-made dataset. It shows that the proposed BC-MF approach is significantly better than the existing approaches, in terms of model accuracy, comprehension of faults, and so on.

Ultra-dense low earth orbit (LEO) satellite network (UD-LSN) is one of the most promising architectures in the sixth-generation (6G) systems, providing several types of services with different service level agreements (SLAs). Network slicing technology effectively meets these SLAs by building multiple logical networks isolated from each other on the physical network. In the UD-LSN, due to the spatiotemporal variations of users and available satellites, it poses a considerable challenge to make dynamic slicing decisions individually for each LEO satellite. This paper proposes a two-layer dynamic reconfigurable radio access network (RAN) slicing architecture for the UD-LSN. We consider the characteristics of enhanced mobile broadband (eMBB) and ultra-reliable and low-latency communications (uRLLC) services and formulate a stochastic optimization problem to maximize the long-term slicing utility, which consists of resource utilization, throughput, and reconfiguration cost. The original problem is transformed into a Markov Decision Process (MDP) and solved with the Branch Dueling Q-Network (BDQ)-based dynamic reconfigurable RAN slicing (DRRS) algorithm in a large slicing window and the priority-based user access algorithm in a small time slot. The simulation results validate the effectiveness of the proposed two-layer DRRS strategy, which has a better performance in the slicing utility, resource utilization, and throughput.

With the advent of edge computing, more and more tasks are offloaded to edge servers, but the computing and storage capabilities of edge servers are limited. Although some works propose efficient schemes for task allocation and pricing, they may ignore users' preferences for continuous tasks. However, the combinatorial preference causes high computational complexity. In this paper, we propose a dominant-strategy incentive compatibility (DSIC) and computationally efficient mechanism for green continuous task allocation based on the combinatorial auction. Besides, the activity on edge (AOE) network is introduced to describe the continuity of tasks. The proposed mechanism gives an approximate solution to the winner determination problem (WDP) in polynomial time and a pricing strategy that can guarantee the truthfulness and individual rationality of auction participants. We demonstrate the approximate ratio of the proposed algorithm through theoretical analysis. Experimental results show that the proposed mechanism achieves truthfulness, individual rationality, and high computational efficiency while considering green continuous task allocation.

Federated learning can collaboratively train AI models while protecting data privacy. In practical industry environment, non-independent and identically distributed (Non-IID) characteristics of data affect the effectiveness of federated learning. Personalized federated learning can help resolve this, but it cannot adapt to unknown data. In addition, practical applications also call for trusted training environment and remain stable when there are security threats. In this article, we propose a credible federated self-learning (CFSL), based on the idea of hypernetwork supported by blockchain to achieve secured, credible, personalized federated self-learning, especially, for unknown data in Non-IID environment. Extensive experiments on three Non-IID data sets demonstrate the capabilities on adaptive resilience for security attacks and on accuracy of recognizing unknown objects, with good performance at the same time. CFSL outperforms the existing personalized federated learning methods, with an increase in average accuracy by 4.11%.

BACKGROUND: Drug-Drug interactions (DDIs) are a challenging problem in drug research. Drug combination therapy is an effective solution to treat diseases, but it can also cause serious side effects. Therefore, DDIs prediction is critical in pharmacology. Recently, researchers have been using deep learning techniques to predict DDIs. However, these methods only consider single information of the drug and have shortcomings in robustness and scalability. RESULTS: In this paper, we propose a multi-type feature fusion based on graph neural network model (MFFGNN) for DDI prediction, which can effectively fuse the topological information in molecular graphs, the interaction information between drugs and the local chemical context in SMILES sequences. In MFFGNN, to fully learn the topological information of drugs, we propose a novel feature extraction module to capture the global features for the molecular graph and the local features for each atom of the molecular graph. In addition, in the multi-type feature fusion module, we use the gating mechanism in each graph convolution layer to solve the over-smoothing problem during information delivery. We perform extensive experiments on multiple real datasets. The results show that MFFGNN outperforms some state-of-the-art models for DDI prediction. Moreover, the cross-dataset experiment results further show that MFFGNN has good generalization performance. CONCLUSIONS: Our proposed model can efficiently integrate the information from SMILES sequences, molecular graphs and drug-drug interaction networks. We find that a multi-type feature fusion model can accurately predict DDIs. It may contribute to discovering novel DDIs.

BACKGROUND: Protein-protein interactions are widespread in biological systems and play an important role in cell biology. Since traditional laboratory-based methods have some drawbacks, such as time-consuming, money-consuming, etc., a large number of methods based on deep learning have emerged. However, these methods do not take into account the long-distance dependency information between each two amino acids in sequence. In addition, most existing models based on graph neural networks only aggregate the first-order neighbors in protein-protein interaction (PPI) network. Although multi-order neighbor information can be aggregated by increasing the number of layers of neural network, it is easy to cause over-fitting. So, it is necessary to design a network that can capture long distance dependency information between amino acids in the sequence and can directly capture multi-order neighbor information in protein-protein interaction network. RESULTS: In this study, we propose a multi-hop neural network (LDMGNN) model combining long distance dependency information to predict the multi-label protein-protein interactions. In the LDMGNN model, we design the protein amino acid sequence encoding (PAASE) module with the multi-head self-attention Transformer block to extract the features of amino acid sequences by calculating the interdependence between every two amino acids. And expand the receptive field in space by constructing a two-hop protein-protein interaction (THPPI) network. We combine PPI network and THPPI network with amino acid sequence features respectively, then input them into two identical GIN blocks at the same time to obtain two embeddings. Next, the two embeddings are fused and input to the classifier for predict multi-label protein-protein interactions. Compared with other state-of-the-art methods, LDMGNN shows the best performance on both the SHS27K and SHS148k datasets. Ablation experiments show that the PAASE module and the construction of THPPI network are feasible and effective. CONCLUSIONS: In general terms, our proposed LDMGNN model has achieved satisfactory results in the prediction of multi-label protein-protein interactions.

Abstract Recently, graph convolutional networks have shown excellent results in skeleton‐based action recognition. This paper presents a multi‐stage part‐aware graph convolutional network for the problems of model over complication, parameter redundancy and lack of long‐dependence feature information. The structure of this network has a multi‐stream input and two‐stream output, which can greatly reduce the complexity and improve the accuracy of the model without losing sequence information. The two branches of the network have the same backbone, which includes 6 multi‐order feature extraction blocks and 3 temporal attention calibration blocks, and the outputs of the two branches are fused together. In multi‐order feature extraction block, a channel‐spatial attention mechanism and a graph condensation module are proposed, which can extract more distinguishable feature and identify the relationship between parts. In temporal attention calibration block, the temporal dependencies between frames in the skeleton sequence are modeled. Experimental results show that the proposed network outperforms many mainstream methods on NTU and Kinetics datasets, for example, it achieves 92.4% accuracy on the cross‐subject benchmark of NTU‐RGBD60 dataset.