IoT device identification stays accurate despite network packet problems
Robust Lightweight IoT Device Fingerprinting Under Network Traffic Anomalies
Networking and Internet ArchitectureMachine Learning
Summary
It can be hard to recognize different internet-connected devices because their network traffic can get jumbled or lost. The authors created a new way to identify these devices quickly by looking only at simple information in the network packets, without needing to see the actual data inside. Their method still works well even if some packets get dropped, mixed up, or copied, and it works on short amounts of traffic. This helps keep networks safe while being efficient enough for devices with limited resources.
What this means in practice
- •For network security teams: Improve device identification reliability in networks even when packet traffic is irregular or corrupted.
- •For iot device manufacturers: Implement scalable device authentication methods that run efficiently on devices with limited compute power to enhance security.
Authors
Roya Taheri, Shamik Sengupta
Abstract
Accurate IoT device identification is essential for network security, yet many existing approaches rely on payload inspection, feature engineering, or long observation windows that limit scalability and often assume stable network conditions despite the inherent variability of real-world IoT traffic. We propose a lightweight IoT device identification approach based on packet headers and locality-sensitive hashing (LSH), generating compact fingerprint representations at packet- and window-level granularities. To evaluate robustness, we consider packet-level traffic anomalies including drop, perturbation, reordering, duplication, and packet injection. Experimental results show that the proposed approach achieves up to 99% accuracy while maintaining robustness under realistic traffic distortions using short bounded observation windows. Anomalies affecting global traffic structure produce the largest performance degradation, whereas window-level representations improve robustness by reducing sensitivity to localized distortions. The low computational overhead of the proposed approach enables scalable IoT device identification in resource-constrained environments.