Radio Frequency Fingerprinting for Frequency Hopping Emitter Identification

In a frequency hopping spread spectrum (FHSS) network, the pattern plays an important role in user authentication at physical layer. However, recently, it has been possible to trace through blind estimation method for (FH) signals. If can be reproduced, attacker imitate FH signal and send fake data FHSS system. To prevent this situation, non-replicable system that targets layer of network is required. study, radio fingerprinting-based emitter identification targeting signals was proposed. A fingerprint (SF) extracted transformed into spectrogram representing time–frequency behavior SF. This trained on deep inception network-based classifier, ensemble approach utilizing multimodality SFs applied. detection algorithm applied output vectors classifier detection. The results showed SF effectively utilized identify with 97% accuracy, detect area under receiver operating characteristic curve 0.99.