Personalized Federated Learning (pFL) tailors models to each client’s local data distribution in heterogeneous federated learning settings. Federated Meta-Learning (FML) is a branch of pFL that uses meta-learning to achieve fast adaptation, where clients start with a meta-model and personalize it by fine-tuning it with local data. Since a single global meta-model has limitations when the data distribution of clients varies significantly, meta-model personalization should be considered in FML. However, most benchmark pFL methods lack meta-model personalization, and usually lack meta-learning or relying on a single global meta-model. Besides, these methods can neither provide meta-model personalization nor guarantee generalization and convergence, due to the challenges in measuring the distance between the meta-model and the client model in FML. To address these issues, we combine FML with hypernetwork and propose a constrained hypernetwork-based FML framework called FMLH, which innovatively utilizes hypernetwork to capture the differences in fine-tuned models, thereby providing personalized meta-models for each client. We provide rigorous mathematical proofs illustrating how the hypernetwork affects the convergence and generalization bounds of FMLH. Experimental results demonstrate that FMLH significantly improves the generalization of the model in cross-client shifts, with the lowest decile accuracy improved by up to 18.71%. FMLH also outperforms representative pFL algorithms by up to 5.6% in terms of maximum accuracy improvement.

Advanced Persistent Threat (APT) attacks—notable for their stealth, sophistication, and potential for catastrophic damage—represent one of the most critical challenges in modern cybersecurity. However, existing detection methods relying on conventional graph models are constrained to capturing pairwise interactions between entities. This limitation hinders the accurate representation of multi-entity dependencies and multidimensional relationships inherent in APT attacks, resulting in imprecise relationship modeling that undermines detection accuracy. Furthermore, efforts to enhance relational representation through complex graph frameworks often introduce significant computational burdens, hindering real-time detection capabilities. To overcome these limitations, we introduce LOHA(Louvain-Hypergraph Masked Autoencoder), a hypergraph modelingdriven APT detection framework that leverages the Louvain algorithm to build expressive hypergraph structures with minimal computational overhead, combined with a mask autoencoder and self-supervised learning for deep semantic feature extraction to enable multi-granularity APT detection. Experimental results demonstrate that LOHA outperforms traditional methods in detection performance across three public datasets while maintaining lower computational overhead, exhibiting enhanced robustness and real-time performance.

The rapid evolution of internet technologies has triggered exponential growth in multimodal data, intensifying security and efficiency challenges in digital copyright protection. This paper proposes a Multimodal Semantic Sharding Graph Convolutional Blockchain System (MSSGC) that innovatively integrates multimodal semantic fragmentation with graph convolutional neural networks (GCN). The system enables load-balanced blockchain dynamic sharding through GCN-based joint clustering of semantic features and transaction networks, effectively reducing cross-shard communication overhead. We develop a Trust-enhanced Proof of Stake (T-PoS) protocol to optimize account sharding via incentive mechanisms while maintaining decentralization and network equilibrium, complemented by queuing theory-based analysis of transactional performance ceilings. Experiments on blockchain simulators demonstrate that MSSGC significantly outperforms baseline systems across diverse sharding configurations: throughput improves by approximately$\mathbf{2 0 \%}$, latency is reduced by$\mathbf{3 0 \%}$, and load-balancing efficiency increases by roughly 2.3. Notably, the system maintains about 70% transaction integrity even under malicious attacks. This work bridges theoretical gaps in semantic-aware copyright protection while advancing practical blockchain implementations for digital rights management.

Federated learning is vulnerable to model poisoning attacks in which malicious participants compromise the global model by altering the model updates. Current defense strategies are divided into three types: aggregation-based methods, validation dataset-based methods, and update distance-based methods. However, these techniques often neglect the challenges posed by device heterogeneity and asynchronous communication. Even upon identifying malicious clients, the global model may already be significantly damaged, requiring effective recovery strategies to reduce the attacker's impact. Current recovery methods, which are based on historical update records, are limited in environments with device heterogeneity and asynchronous communication. To address these problems, we introduce FedHAN, a reliable federated learning algorithm designed for asynchronous communication and device heterogeneity. FedHAN customizes sparse models, uses historical client updates to impute missing parameters in sparse updates, dynamically assigns adaptive weights, and combines update deviation detection with update prediction-based model recovery. Theoretical analysis indicates that FedHAN achieves favorable convergence despite unbounded staleness and effectively discriminates between benign and malicious clients. Experiments reveal that FedHAN, compared to leading methods, increases the accuracy of the model by 7.86%, improves the detection accuracy of poisoning attacks by 12%, and enhances the recovery accuracy by 7.26%. As evidenced by these results, FedHAN exhibits enhanced reliability and robustness in intricate and dynamic federated learning scenarios.

Against the backdrop of the rapid proliferation of Industrial Internet of Things, due to the rapid increase in edge devices and multi-modal data, Multi-modal Federated Learning (MFL)—an extension of Federated Learning (FL)—has become the mainstream solution for fusing heterogeneous sensor data in edge computing. Although MFL is inherently vulnerable to backdoor attacks similar to FL, its cross-modal fusion techniques for verifying semantic consistency gradually may cause single-modal triggers to be faded during training. To address the aforementioned issue, we propose a cross-modality bi-trigger backdoor attack against MFL, named Bad-MFL, which is the first backdoor attack specifically targeting MFL. Bad-MFL employs two trigger generation modes to randomly implant logically correlated, invisible bi-trigger in two modalities. By maintaining semantic consistency between the triggers, Bad-MFL bypasses cross-modal fusion validation and successfully implants a backdoor into the global model. Moreover, it ensures the backdoor will not disappear as training progresses. Our experiments indicate that, compared to traditional backdoor attack methods, Bad-MFL achieving an Attack Success Rate up to 89.04%, which is 56% higher than the baseline attack, and its backdoor remains effective in high heterogeneous environments throughout training.

In autonomous driving environments, generative adversarial networks (GANs) are often used to predict the future trajectories of objects in the scene, providing decision support for autonomous driving systems. However, integrating GAN models into the Internet of Vehicles (IoV) poses numerous challenges. First, GAN models necessitate user data and extensive computing resources, whereas diverse intelligent connected vehicle (ICV) possess limited bandwidth and computational capabilities, making it challenging to deploy models of the same scale as those in the cloud. Second, multifaceted aspects, including energy consumption, computation, communication, and vehicle training scheduling, have yet to be thoroughly examined, particularly in the context of IoV’s limited resources. To address the above issues, we propose a novel federated learning framework, heterogeneous-vehicle-scheduling-GAN (HVS-GAN), for training GANs in resource-constrained IoV environments. HVS-GAN balances GAN generation quality and training costs in IoV. It supports multiple ICVs training GAN models of different structures, breaking the strong assumption of uniform GAN model size constraints in previous works and enabling collaborative learning within IoV. Furthermore, to balance quality and training costs, we incorporate deep deterministic policy gradients learning to manage varying model size constraints, training delays, and training consumption across participating ICVs. Experimental results and analysis confirm the superiority of our proposed HVS-GAN solution, which achieves better outcomes in IoV scenarios with stringent model size constraints compared to state-of-the-art algorithms.

Dataset-Based Defense is a common proactive defense method aimed at countering backdoor attacks by filtering and removing poisoned samples from the training dataset. Recent research has utilized diffusion models to purify contaminated samples. However, during both forward and backward denoising processes, the inherent noise and distribution characteristics of the data can also affect the purification of poisoned samples, making the process unreliable. To address this issue, this paper re-examines the purification process from the perspective of causal graphs, revealing how backdoor triggers and noise act as confounding factors, generating false associations between contaminated images and labels. To eliminate this false association, we propose a Causality-Augmented Diffusion Model-Based Defense (CADMBD) method. CADMBD applies multiple iterations of forward noise addition and backward denoising to the data input in the diffusion model. During the backward denoising process, residual calculation and causal layer adjustment are used to eliminate backdoor pathways, thereby establishing a true causal relationship between benign features and labels. Finally, the most suitable generated image is selected as the final purified image based on pixel and feature distances from each iteration. Experimental results demonstrate that CADMBD effectively addresses various complex attacks, significantly enhancing the purification effect while minimizing the impact on benign sample accuracy, outperforming baseline defense methods.