MFCC-Based Radio Frequency Fingerprint Extraction and Emitter Recognition

Driven by advances in wireless communication technology, the need for precise device identification and authentication has become increasingly urgent. As a device identification technology based on the physical layer, Radio Frequency Fingerprint (RFF) has significant research value. The Mel-Frequency Cepstral Coefficients (MFCC), as a mature voice feature extraction method, offers a novel approach to RFF with its unique frequency domain analysis and feature extraction capabilities. This study proposes a novel feature extraction method based on MFCC from air interface signals, and employs platforms using Support Vector Machine (SVM) and Convolutional Neural Network (CNN) for feature classification and recognition. The experimental results demonstrate that the CNN-based approach achieves a recognition rate of up to 94.4% under distinguishing 84 modules. Overall, the application of MFCC in RFF identification presents a novel and effective solution to address the challenges in authentication and identification of specific emitter.