Edge AI-based Radio Frequency Fingerprinting for IoT Networks

Download PDF | BibTeX

Abstract: The deployment of the Internet of Things (IoT) in smart cities and critical infrastructure has enhanced connectivity and real-time data exchange but introduced significant security challenges. While effective, cryptography can often be resource-intensive for small-footprint, resource-constrained IoT devices. Radio Frequency Fingerprinting (RFF) offers a promising alternative by leveraging unique RF signal characteristics for device identification at the Physical (PHY) layer. This paper proposes lightweight AI-based RFF schemes tailored for edge devices, demonstrating their deployment feasibility on resource-constrained environments like Raspberry Pi. Two models are presented—a Convolutional Neural Network (CNN) and a Transformer Encoder—both optimized for Edge AI using TensorFlow Lite. Evaluation results reveal high classification accuracy (>95%) and ROC-AUC (>90%), with minimal model sizes suitable for IoT applications.