Specific Emitter Identification Based on Software-Defined Radio and Decision Fusion

Kaiwen Tan,Wenjun Yan,Limin Zhang,Yan Zhang,Miao Tang

doi:10.1109/access.2021.3088542

Kaiwen Tan, Wenjun Yan + Show 3 more

Open Access

https://doi.org/10.1109/access.2021.3088542

Copy DOI

Abstract

Specific emitter identification (SEI) uses the unintentional modulation information carried by the emitter waveform, i.e., radio frequency fingerprints, to realize the matching identification of the received signal and its corresponding emitter. We propose an emitter identification scheme based on deep learning (DL). The received signal is subjected to time-varying filtered empirical mode decomposition (tvf-EMD). The obtained intrinsic mode functions (IMFs) are subjected to Hilbert transformation to obtain a 3D-Hilbert spectrum, using amplitude frequency aggregation characteristics obtained by projection to express the nonstationary information of the signal and calculate its bispectral diagonal slice as the second feature. A squeeze-and-excitation block (SEB) is introduced to a convolutional neural network (CNN) to focus on the effective area in the feature map, and support vector machine (SVM) is used to fuse the decision information of the two types of features. We collect four types of steady-state signals on a software-defined radio (SDR) sensor platform based on GNU Radio and universal software radio peripherals (USRP), and study the algorithm performance. The results show that our scheme has strong generalization, high recognition accuracy, and broad application prospects compared with other SEI methods.

Highlights

Specific emitter identification (SEI) uses feature parameters to describe emitter signals and identify individual radio emitters
We present an SEI scheme based on a deep learning (DL) framework
We use time-varying filtered empirical mode decomposition, and reduce the dimension of the Hilbert spectrum. We believe this is the first use of intrinsic mode functions (IMFs) obtained from tvf-EMD in SEI

Summary

INTRODUCTION

Specific emitter identification (SEI) uses feature parameters to describe emitter signals and identify individual radio emitters. We use time-varying filtered empirical mode decomposition (tvf-EMD), and reduce the dimension of the Hilbert spectrum We believe this is the first use of IMFs obtained from tvf-EMD in SEI. The local cutoff frequency of each segment was obtained separately, and these were used as IMFs. In SEI, the IMFs of tvf -EMD are used for a Hilbert transform, aiming to improve the frequency separation performance and extract the additional noise of the hardware. We project the Hilbert spectrum processed by tvf-EMD into the amplitude frequency domain, and the difference features are extracted by using the robustness of the signal in the time domain. It can be seen that the low frequency components of different emitters have high similarity, the difference is mainly reflected in the medium- and highfrequency components, and the time-series features of the same emitter are similar

FEATURE EXTRACTION BASED ON BISPECTRUM

Squeeze-and-Excitation Block

CONCLUSION

Findings

Computational Complexity