Preamble-based detection of Wi-Fi transmitter RF fingerprints

Yuan, Honglin; Hu, Aiqun

doi:10.1049/el.2010.1220

Cited by 46 publications

(24 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In other words, we have four classes of data, which are denoted as T i , i = 1, 2, 3, 4, and there are 40 bursts in each class. After obtaining the data, we then extract the transient part of each burst by using a method similar to that in [28] and normalize these transients to [0,1]. Since the number of bursts is small and the transient part of each burst is very short (about 0.83 milliseconds), we will do the experiment just as the sampling with replacement.…”

Section: Resultsmentioning

confidence: 99%

Specific Emitter Identification Based on the Natural Measure

Jia

Zhu

Gan

2017

Entropy

View full text Add to dashboard Cite

Specific emitter identification (SEI) techniques are often used in civilian and military spectrum-management operations, and they are also applied to support the security and authentication of wireless communication. In this letter, a new SEI method based on the natural measure of the one-dimensional component of the chaotic system is proposed. We find that the natural measures of the one-dimensional components of higher dimensional systems exist and that they are quite diverse for different systems. Based on this principle, the natural measure is used as an RF fingerprint in this letter. The natural measure can solve the problems caused by a small amount of data and a low sample rate. The Kullback-Leibler divergence is used to quantify the difference between the natural measures obtained from diverse emitters and classify them. The data obtained from real application are exploited to test the validity of the proposed method. Experimental results show that the proposed method is not only easy to operate, but also quite effective, even though the amount of data is small and the sample rate is low.

show abstract

Section: Resultsmentioning

confidence: 99%

Specific Emitter Identification Based on the Natural Measure

Jia

Zhu

Gan

2017

Entropy

View full text Add to dashboard Cite

show abstract

“…The methods used to analyse the steady-state signal are a lot diverse and include usage of modulation-based approaches [4], wavelet-based approaches [5], cyclostationary-based approaches, preamble-based approaches [6]. One of the popular wavelet-based approach, termed as the dynamic wavelet fingerprint approach, utilises wavelet transforms on the steady signal in its time domain [5].…”

Section: Related Workmentioning

confidence: 99%

“…One of the popular wavelet-based approach, termed as the dynamic wavelet fingerprint approach, utilises wavelet transforms on the steady signal in its time domain [5]. The features generated in the modulation-domain calculate the error between the ideal demodulated signal and the transmitted signal [4], whereas the preamble-based approach analyse the features of the preamble extracted, namely its periodicity [6]. The cyclostationary-based approach, analyse the salient cyclic features present in a signal.…”

Section: Related Workmentioning

confidence: 99%

Analysis of RF Device Fingerprinting using Convolutional Neural Network

Parvathi¹,

Basu²

2020

IJERT

View full text Add to dashboard Cite

The authentication of devices and identification of unauthorised signals has been difficult in many scenarios. In this paper, a method for identifying a specific device among a range of similar devices is discussed using the techniques of convolutional neural network. The approach is to use radio fingerprints as samples for the identification process. The variability caused in the manufacture of these wireless transmitters serves as a unique feature, creating repeatable signatures for a particular device. These signatures are used as fingerprints for identification and verification. A framework is then developed for training the convolutional neural network using complex baseband error signals in time domain for varying signal-to-noise ratios (SNRs).

show abstract

“…The accuracy of the detection and separation of the turnon transients is crucial to the overall identification performance. The Bayesian ramp detector, variance trajectory detector and correlation detector have been reported for the detection of the start instants of transients [7][8][9]. The RF fingerprint features are generated from instantaneous amplitude, phase, and frequency responses of transient signal [10].…”

Section: Introductionmentioning

confidence: 99%

“…However, the discriminatory performance of transient analysis is not always good when the emitters are the same model [12][13]. The underlying reason is that the detection of the exact turn-on time instant of transient is difficult because of the unstable profiles of transient [9]. Another reason which makes transient analysis impractical may be that it requires high-end receiver to offer high oversampling rate [14].…”

Section: Introductionmentioning

confidence: 99%