Federated learning is a distributed machine learning approach that enables multiple participants to collaboratively train a model without sharing their data, thus preserving privacy. However, the decentralized nature of federated learning also makes it susceptible to backdoor attacks, where malicious participants can embed hidden vulnerabilities within the model. Addressing these threats efficiently and effectively is crucial, especially given the impracticality of iterative and resource-intensive detection methods in federated learning environments. This paper presents a novel framework for one-shot backdoor removal in federated learning. Our approach integrates advanced anomaly detection techniques with a unique model update aggregation strategy, allowing for the identification and neutralization of backdoor influences in a single update cycle without the need for extensive data access or communication between participants. Extensive experiments across various federated architectures and data distributions demonstrate that our method effectively mitigates backdoor threats while maintaining model performance and scalability. This work not only enhances the security of federated models but also contributes to the broader applicability of federated learning in sensitive and critical domains.

This paper presents a novel and robust watermarking framework for diffusion models in federated learning (FL) environments, designed to safeguard intellectual property through unlearning-enhanced redundancy. Unlike prior methods that apply watermarks after model training, our approach embeds structured, imperceptible watermarks directly into model parameters during the federated optimization process. We introduce a dual-purpose loss function that jointly drives generative training and watermark embedding, maintaining model quality while enabling reliable traceability. To enhance watermark robustness, we propose an unlearning-guided reallocation strategy that leverages machine unlearning dynamics to adaptively reposition the watermark in stable parameter subspaces. This embedding process is mathematically modeled via a stochastic differential equation, capturing the uncertainties inherent to FL. Additionally, we parameterize the watermark path over a complex manifold using series expansions to control its spatial and spectral characteristics. Spectral analysis is then applied to further improve resilience against model pruning, quantization, and fine-tuning attacks. Extensive experiments on MNIST, CIFAR-10, and CIFAR-100 under both IID and non-IID data splits demonstrate superior average watermark accuracy (AWA) and watermark efficiency (WE) across varied client settings. For instance, with five non-IID clients on MNIST, our method achieves 98.5% AWA and 78.5% WE, outperforming existing techniques. These results underscore the effectiveness of unlearning-enhanced watermark redundancy as a scalable and secure solution for ownership protection in distributed diffusion models.

The advent of 6G networks brings forth unprecedented security challenges due to massive device interconnectivity and rapidly evolving cyber threats. Existing centralized anomaly detection and defense systems fall short in scalability, privacy protection, and responsiveness, rendering them ineffective for next-generation network environments. In this work, we propose FedGAD (Federated Generative Anomaly Detection), a technically novel framework that tightly integrates federated learning with generative adversarial training to enable distributed, privacy-preserving, and adaptive anomaly detection. Unlike prior approaches, FedGAD introduces a dual-model architecture where a federated generator synthesizes hard-to-detect adversarial anomalies, while local discriminators are collaboratively trained across edge nodes to improve detection sensitivity without sharing raw data. This adversarial co-training process enhances model robustness against zero-day attacks and adapts in real-time to changing threat landscapes. Extensive experiments show that FedGAD outperforms state-of-the-art techniques in terms of detection accuracy, adaptability, and computational efficiency, offering a scalable and resilient defense paradigm tailored for the dynamic security demands of 6G networks.

Federated learning (FL) has gained significant traction as a paradigm for decentralized learning, enabling multiple clients to collaboratively train models without sharing their local data. However, applying FL to graph-structured data introduces unique challenges, such as handling non-IID data and preserving the structural dependencies between nodes. Additionally, existing approaches to FL with graph neural networks (GNNs) often struggle to capture complex relationships within graph data and are vulnerable to privacy breaches, including membership inference and model inversion attacks. In this article, we propose federated capsule graph neural networks (FCGNNs), a novel architecture that integrates the dynamic routing capabilities of capsule networks with the structure-preserving power of GNNs in a federated setting. FCGNN is designed to effectively model hierarchical and part-whole relationships within graph data, enabling it to outperform traditional Federated GNN approaches. We enhance the privacy of FCGNN by incorporating differential privacy and secure aggregation techniques, ensuring that individual client updates remain confidential while maintaining strong model performance. We evaluate FCGNN on several benchmark graph datasets, including Cora, Citeseer, PubMed, and PROTEINS, and demonstrate that it consistently achieves higher accuracy and F1-scores compared to existing FL methods. Our experiments show that FCGNN converges faster and incurs lower communication costs, making it highly efficient for real-world FL applications. Furthermore, FCGNN is robust across different numbers of participating clients, maintaining high performance even in non-IID scenarios. These results highlight the potential of FCGNN as a scalable and privacy-preserving solution for decentralized learning on graph-structured data.

Personalized lesson planning is time-consuming and demands simultaneous alignment to curriculum standards, classroom constraints, and individual teacher style. Although large language models (LLMs) can draft lesson plans, their outputs often remain generic, lack verifiable grounding in official curricula, and require substantial teacher revision to become classroom-ready. We present TeachPlanAlign, a dual-profile lesson plan generation framework that (i) models both teacher preferences and class learning context, (ii) grounds generation in curriculum evidence through retrieval-augmented fine-tuning designed for open-book generation, and (iii) iteratively improves pedagogical coherence via a constraint-guided self-refinement loop. The framework produces structured plans with explicit time allocation, differentiation strategies, and an evidence-linked rationale that supports traceability to standards and instructional resources. We evaluate TeachPlanAlign on a multi-subject benchmark of lesson requests paired with curriculum documents and human-authored plans, and we further validate its usability through teacher-in-the-loop evaluation. Results show consistent improvements in curriculum alignment, evidence faithfulness, and teacher preference satisfaction, while reducing teacher editing effort. These findings suggest that dual-profile alignment with traceable curriculum grounding can improve the usability and auditability of LLM-based lesson planning on this benchmark, but they should not yet be interpreted as evidence of deployment readiness.