The deployment of Internet of Things (IoT) devices requires efficient security mechanisms. However, cryptographic solutions often prove resource-intensive. Radio Frequency Fingerprinting (RFF) enables device authentication through the intrinsic characteristics of RF signals at the Physical (PHY)-layer. Deploying RFF presents two challenges: ensuring operational efficiency and scalability in resource -constrained environments. This paper presents a lightweight Edge AI -based RFF model for device authentication using PHY-layer characteristics. Our approach implements a Deep Learning (DL) model to extract device-specific features from IQ samples, converted using TensorFlow Lite for edge deployment. Evaluation on Raspberry Pi demonstrates high accuracy (> 0.95) and Receiver Operating Characteristic-Area Under the Curve (ROC-AUC) scores (> 0.90), while maintaining a compact model size suitable for resource-constrained environments.