Award-Winning Research on Real-Time Packet-Based Intrusion Detection on Edge Devices

We are excited to announce that the groundbreaking paper by Borgioli Niccolò, Linh Thi Xuan Phan, Federico Aromolo, Alessandro Biondi, and Giorgio Buttazzo, from the ReTiS laboratory of Scuola Superiore Sant’Anna, along with Linh Thi Xuan Phan from the University of Pennsylvania, has received the prestigious Best Paper Award at the RAGE workshop during 2023 Cyber-Physical Systems and IoT Week (CPS-IoT-Week 2023). Their paper titled “Real-Time Packet-Based Intrusion Detection on Edge Devices” has been evaluated by a panel of international experts, who deemed it the best contribution to the workshop.

In today’s increasingly connected world, cyber-physical systems face a growing number of security threats that are becoming more advanced and complex. Traditional security systems called Intrusion Detection Systems (IDSs) rely on predefined patterns to identify attacks. However, they struggle to keep up with the constantly evolving attack techniques, making them less effective in detecting novel and sophisticated attacks. This award-winning research introduces an innovative Intrusion Detection System based on autoencoder technology. This intelligent system is capable of learning the expected patterns of network traffic and effectively detecting not only known attacks but also previously unseen ones. By using advanced machine learning algorithms, the autoencoder-based IDS demonstrates exceptional precision in identifying new malicious packets. What sets this research apart is its versatility. The proposed technique can detect a wide range of attacks, making it highly adaptable and applicable in various scenarios. The team conducted extensive experiments using simulations and real hardware to showcase the superior performance of their approach compared to existing solutions in terms of detection accuracy and generality. Additionally, the researchers analyzed the speed at which the detection mechanism operates, providing valuable insights into its predictability and practical usability. This is particularly important for resource-constrained edge devices, where real-time intrusion detection is vital to ensuring the security of sensitive systems. The recognition received by Borgioli Niccolò and the team highlights the ReTiS laboratory’s commitment to pushing the boundaries of research in the field of cyber-physical systems and security. The collaboration with Linh Thi Xuan Phan from the University of Pennsylvania emphasizes the significance of interdisciplinary cooperation in addressing complex challenges. The Best Paper Award acknowledges the exceptional quality and impact of this research. We extend our warmest congratulations to the entire team for their remarkable achievement. For more information about this award-winning research and the CPS IoT Week, please visit the conference website.