Specific Emitter Identification Based on Deep Residual Networks

Abstract

Specific emitter identification (SEI) enables the discrimination of individual radio emitters with the external features carried by the received waveforms. This identification technique has been widely adopted in military and civil applications. However, many previous methods based on hand-crafted features are subject to the present expertise. To remedy these shortcomings, this paper presents a novel SEI algorithm using deep learning architecture. First, we perform Hilbert-Huang transform on the received signal and convert the resulting Hilbert spectrum into a grayscale image. As a signal representation, the Hilbert spectrum image has high information integrity and can provide abundant information about the nonlinear and non-stationary characteristics of signals for identifying emitters. Thereafter, we construct a deep residual network for learning the visual differences reflected in the Hilbert spectrum images. By using the residual architectures, we effectively address the degradation problem, which improves efficiency and generalization. From our analysis, the proposed approach combines high information integrity with low complexity, which outperforms previous studies in the literature. The simulation results validate that the Hilbert spectrum image is a successful signal representation, and also demonstrate that the fingerprints extracted from raw images using deep learning are more effective and robust than the expert ones. Furthermore, our method has the capability of adapting to signals collected under various conditions.

Radio Frequency Fingerprint Identification Based on Denoising Autoencoders

“…Although various DL RFFs have been demonstrated to achieve excellent identification in high SNR regions, their performance in low SNRs is far from satisfactory. For instance, at SNR = 10 dB, the classification accuracies were only 73.73% for seven ZigBee devices [7], 38% for 12 mobile phones [9], 58% for seven USRPs [13], and 65% for five simulated radio emitters [12], respectively. However, since most of the IoT devices are battery-powered, the transmitting power is relatively low, they usually need to work in low SNR scenarios.…”

Generative Adversarial Network Based Rogue Device Identification Using Differential Constellation Trace Figure

“…With the help of the rapid development of deep learning technologies, deep learning-based methods have been widely used in the field of wireless communication, such as channel estimations, waveform angle estimations, and modulation type identifications [24]. Many studies have used them for RF fingerprint-based device classification problems [16,[25][26][27][28][29][30]. In our previous work, a DCTF based convolutional neural network (CNN) system was designed to classify 54 Zigbee devices [30].…”

“…Kose et al used probabilistic neural network (PNN) to classify the RF fingerprints extracted from transient signals of WiFi devices [28]. Pan et al used deep residual networks to train the Hilbert spectrum images of received signals to classify specific emitters [29].…”

Location-Invariant Physical Layer Identification Approach for WiFi Devices

“…PLS exploits the inherent features of devices and wireless channels to authenticate users and encrypt the communication data [5]. As one of the PLS technologies, Radio Frequency Fingerprinting (RFF) is adopted herein as a way to augment existing multifactor authentication schemes at the device level to counter forgery and related threats [6]- [9]. RFF identifies a transmitter through discriminating features (also called patterns) extracted from its intrinsic physical properties [10].…”