The field of Internet of Things (IoT) security and federated learning is rapidly evolving, with a focus on developing innovative solutions to address the challenges of privacy, scalability, and communication efficiency. Recent studies have explored the use of streaming learning approaches, such as adaptive random forests and Hoeffding adaptive trees, to detect anomalies in IoT traffic and improve robustness. Additionally, researchers have proposed novel frameworks for enhancing random forest classifiers and federated learning models, including the use of simulated annealing and mist-assisted hierarchical architectures. These advancements have shown promising results in improving predictive accuracy, generalization, and fairness in client participation. Noteworthy papers include: Mist-Assisted Federated Learning for Intrusion Detection in Heterogeneous IoT Networks, which proposes a hierarchical framework for IoT intrusion detection and achieves high accuracy and stable convergence. Federated Quantum Kernel Learning for Anomaly Detection in Multivariate IoT Time-Series, which integrates quantum feature maps with federated aggregation for distributed anomaly detection and shows superior generalization in capturing complex temporal correlations.