Feature Correlation Based RF Fingerprinting Identification for Walkie-Talkies Push-to-Talk Communications

Abstract

Walkie-talkies play a key role in public safety and military communications. To counter the security threats posed by counterfeiting and illegal use, this paper proposes a radio frequency (RF) fingerprinting method for walkie-talkies based on interpretable features and multi-layer perceptron (MLP). By extracting interpretable features in time, frequency, and statistical domains affected by different RF impairments and performing correlation analysis and dimensionality reduction, it ensures that the acceptable low complexity requirements of walkie-talkie equipment are maintained while the recognition capability is preserved. The measured results show that the recognition accuracy of the method can reach 96.5%, which has the advantages of being efficient, stable and interpretable. This research provides a low complexity and reliable physical layer security solution for identifying legitimate and illegitimate walkie-talkies, which can be further used to guide the engineering implementation and application of RF fingerprint identification modules.