Specific Emitter Identification via Convolutional Neural Networks

Ding, Lu; Wang, Shilian; Wang, Fanggang; Zhang, Wei

doi:10.1109/lcomm.2018.2871465

Cited by 188 publications

(79 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A CNN was used in [22] for fingerprinting identification of five ZigBee devices. Bispectrum of the received signal was calculated as a unique feature in [33] and a CCN was used to identify specific emitters. A CNN was also used in [34] to identify five USRP devices.…”

Section: A Related Workmentioning

confidence: 99%

Deep Learning for Large-Scale Real-World ACARS and ADS-B Radio Signal Classification

et al. 2019

View full text Add to dashboard Cite

Radio signal classification has a very wide range of applications in the field of wireless communications and electromagnetic spectrum management. In recent years, deep learning has been used to solve the problem of radio signal classification and has achieved good results. However, the radio signal data currently used is very limited in scale. In order to verify the performance of the deep learning-based radio signal classification on real-world radio signal data, in this paper we conduct experiments on largescale real-world ACARS and ADS-B signal data with sample sizes of 900,000 and 13,000,000, respectively, and with categories of 3,143 and 5,157 respectively. We use the same Inception-Residual neural network model structure for ACARS signal classification and ADS-B signal classification to verify the ability of a single basic deep neural network model structure to process different types of radio signals, i.e., communication bursts in ACARS and pulse bursts in ADS-B. We build an experimental system for radio signal deep learning experiments. Experimental results show that the signal classification accuracy of ACARS and ADS-B is 98.1% and 96.3%, respectively. When the signal-to-noise ratio (with injected additive white Gaussian noise) is greater than 9 dB, the classification accuracy is greater than 92%. These experimental results validate the ability of deep learning to classify large-scale real-world radio signals. The results of the transfer learning experiment show that the model trained on large-scale ADS-B datasets is more conducive to the learning and training of new tasks than the model trained on small-scale datasets.INDEX TERMS Radio signal classification, deep learning, convolutional neural network, residual network, Inception, cognitive radio, ACARS, ADS-B.

show abstract

Section: A Related Workmentioning

confidence: 99%

Deep Learning for Large-Scale Real-World ACARS and ADS-B Radio Signal Classification

et al. 2019

View full text Add to dashboard Cite

show abstract

“…As a result, recent published REC work mainly focuses on two aspects, feature extraction and classifier selection. For feature extraction, various time-frequency (T-F) transform approaches, like short-time Fourier transform (STFT) [5], wavelet [6], quadratic T-F [7], Zhang et al [8] and variational mode decomposition (VMD) [9] are adopted to extract more distinguishable intra-pulse features including intentional or even unintentional modulation features [9], [10]. Currently, with the introduction of more complex modulation techniques, the dimensions of feature vectors continue to increase.…”

Section: Introductionmentioning

confidence: 99%

Online Non-Cooperative Radar Emitter Classification From Evolving and Imbalanced Pulse Streams

Sui

Liu

et al. 2020

IEEE Sensors J.

View full text Add to dashboard Cite

Recent research treats radar emitter classification (REC) problems as typical closed-set classification problems, i.e., assuming all radar emitters are cooperative and their pulses can be pre-obtained for training the classifiers. However, such overly ideal assumptions have made it difficult to fit real-world REC problems into such restricted models. In this paper, to achieve online REC in a more realistic way, we convert the online REC problem into dynamically performing subspace clustering on pulse streams. Meanwhile, the pulse streams have evolving and imbalanced properties which are mainly caused by the existence of the non-cooperative emitters. Specifically, a novel data stream clustering (DSC) algorithm, called dynamic improved exemplar-based subspace clustering (DI-ESC), is proposed, which consists of two phases, i.e., initialization and online clustering. First, to achieve subspace clustering on subspace-imbalanced data, a static clustering approach called the improved ESC algorithm (I-ESC) is proposed. Second, based on the subspace clustering results obtained, DI-ESC can process the pulse stream in real-time and can further detect the emitter evolution by the proposed evolution detection strategy. The typically dynamic behavior of emitters such as appearing, disappearing and recurring can be detected and adapted by the DI-ESC. Extinct experiments on real-world emitter data show the sensitivity, effectiveness, and superiority of the proposed I-ESC and DI-ESC algorithms.

show abstract

“…One is the intelligent sorting algorithms such as the clustering algorithm or classifier, which can process the feature-parameter uncertainty or overlapping of the measured emitter signals and classify them correctly. At present, many intelligent sorting algorithms and classifiers have been successfully applied, such as fuzzy clustering [6], K-means clustering [7], grid clustering [8], density clustering [9], support vector machine classification [10], neural networks [11,12], and so on. However, some of them can only solve the problem of the uncertainty or overlapping of the emitter feature parameters to a certain extent, and some of them are very sensitive to signal-to-noise ratios (SNRs).…”

Section: Introductionmentioning

confidence: 99%

Radar Signal Sorting Method Based on Radar Coherent Characteristic

et al. 2020

View full text Add to dashboard Cite

Aiming at the problem of reliability reduction of signal sorting in terms of the traditional five parameters and intrapulse feature in a complex electromagnetic environment, a new signal sorting method based on radar coherent characteristics is proposed. The main idea of this method is using spectrum analysis to obtain the spectrum images of coherent and noncoherent signals. Image-processing technology is used to extract the feature difference between the two spectrum images, and the central-moment feature is introduced to describe this difference. Through simulation analysis, the feasibility of using the central-moment feature as the coherent feature for signal sorting was proved. In order to check the effectiveness of the proposed feature, a number of simulations were conducted to demonstrate the sorting capability in terms of the coherent feature. From the simulations, it can be seen that the proposed feature not only can be used as a new feature for signal sorting but also that it can be utilized as a supplement for five typical parameters and the intrapulse feature to improve the sorting accuracy rate. Simulations also showed the proposed method could achieve satisfactory sorting results in a low signal-to-noise ratio (SNR). When the SNR was 5 dB, the sorting accuracy rate could reach 98%.

show abstract

Specific Emitter Identification via Convolutional Neural Networks

Cited by 188 publications

References 13 publications

Deep Learning for Large-Scale Real-World ACARS and ADS-B Radio Signal Classification

Deep Learning for Large-Scale Real-World ACARS and ADS-B Radio Signal Classification

Online Non-Cooperative Radar Emitter Classification From Evolving and Imbalanced Pulse Streams

Radar Signal Sorting Method Based on Radar Coherent Characteristic

Contact Info

Product

Resources

About