2023 · 10 citations · 12 references
ComputingComputer ScienceElectrical EngineeringFeature EngineeringMachine LearningWireless CommunicationsDeep LearningFeature ExtractionPattern RecognitionRadio-frequency Identification SystemsRepresentation LearningSignal ProcessingBiometric Feature ExtractionFeature LearningFingerprint RecognitionFingerprintingNon-ideal EffectsRadio Frequency FingerprintingRadio-frequency IdentificationTraining
The Radio Frequency Fingerprint Identification (RFFI) technique utilizes the subtle and unintentional modulation present in the transmitted RF waveform, caused by the non-ideal characteristics of the device, to uniquely identify an emitter. RFFI has recently gained significant attention, and deep learning (DL) approach has achieved superior performance. In this paper, we study the DL-based RFFI with an emphasis on the fingerprint contamination by receivers. Specifically, during reception, the receivers’ fingerprints are inevitably introduced, and different receivers affect emitters’ fingerprints differently, thereby degrading recognition performance when the RFFI recognition model is deployed at different receivers. To mitigate the receivers’ effect, we propose a receiver-independent recognition model based on the idea of feature disentanglement. With the aid of some carefully designed model structure, loss functions and training strategy, the proposed model learns a feature space, in which the emitters’ features are disentangled from the receivers’, through multi-receiver’s training data. When deploying the model on a new receiver, the emitter-dependent features are exploited to achieve receiver-independent identification. Experiments on two real-world datasets demonstrate that our method significantly outperforms the baseline method without considering fingerprint contamination.
12
Deep Residual Learning for Image Recognition
Kaiming He, Xiangyu Zhang, Shaoqing Ren et al. · 2016 · 222.1K citations · Full text
Domain Generalization: A Survey
Kaiyang Zhou, Ziwei Liu, Yu Qiao et al. · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2022 · 1.1K citations
Specific Emitter Identification Based on Deep Residual Networks
Yiwei Pan, Sihan Yang, Hua Peng et al. · IEEE Access · 2019 · 165 citations · Full text