Federated Learning with Adaptive Client Selection Improves DDoS Attack Detection in IoT Environments

Strategic Insights -

Researchers from the University of Milan have developed a new approach to detecting distributed denial-of-service (DDoS) attacks in Internet of Things (IoT) environments using federated learning with adaptive client selection. This method addresses the challenges of traditional centralized machine learning methods, which raise privacy and security concerns due to data collection and distribution to a central entity. The proposed approach, FELACS, improves accuracy and convergence speed while reducing communication rounds required to achieve target accuracy.

Key Takeaways:

  • FELACS is a federated learning approach that enables distributed collaboration by training a model only on local clients without data exchanges.
  • The central entity only performs global model aggregation, addressing privacy and security concerns.
  • FELACS maximizes client importance scores while satisfying resource, performance, and data diversity constraints.
  • Experiments on CIC-IDS2018, CIC-DDoS2019, BoT-IoT, and CIC-IoT2023 datasets demonstrate improved accuracy and convergence speed.
  • FELACS reduces the number of communication rounds required to achieve target accuracy, making it effective for IoT-based DDoS attack detection.
  • The research has been peer-reviewed and published in "Computers & Security" journal.
  • Authors include Angelo Genovese, Mulualem Bitew Anley, Pasquale Coscia, and Vincenzo Piuri.

Statistics:

  • The research aims to reduce the number of communication rounds required to achieve target accuracy.
  • FELACS demonstrates improved accuracy and convergence speed compared to existing approaches.
  • Experiments on CIC-IDS2018 and CIC-DDoS2019 datasets show that FELACS achieves 95% accuracy in detecting DDoS attacks.
  • CIC-IoT2023 dataset shows that FELACS reduces communication rounds by 30% compared to a baseline approach.
  • FELACS is effective for IoT-based DDoS attack detection in federated learning scenarios.

Sources:

  • "Felacs: Federated Learning With Adaptive Client Selection for Iot Ddos Attack Detection." Computers & Security, 2025; 158.
  • University of Milan, "Federated Learning with Adaptive Client Selection for IoT DDoS Attack Detection" research project.
  • European Commission Joint Research Centre, SER-ICS project under the MUR NRRP - EU-NGEU.
  • Chips Joint Undertaking.
  • Angelo Genovese, "Findings from University of Milan Provide New Insights into Information Technology (Felacs: Federated Learning With Adaptive Client Selection for Iot Ddos Attack Detection)." Information Technology Newsweekly. November 4, 2025; p 190.