Work

Device Identification Using Discrete Wavelet Transform

Year: 2021
Type: conference-paper
Abstract: This paper investigates the effectiveness of employing measured hardware features mapped into the frequency domain for devices identification. The technique is to utilize Discrete Wavelet Transform (DWT) coefficients as distinguishing features. The DWT coefficients address the degree of relationship between the investigated features and the wavelet function at different occurrences of time. Therefore, DWT coefficients carry useful temporal information about the transient activity of the investigated wavelet features. We study the impacts of utilizing different wavelet functions (Coiflets, Haar and Symlets) on the performance of the device identification system. This system yields 92.5 % of accuracy using Sym6 wavelet. A comparison is made of the accuracy of wavelet features and raw features with standard classifiers.
Authors: Supriya Yadav, Pooja R. Khanna, Gareth Howells
Institution: University of Kent
Language: English
Open Access status: green
OpenAlex ID: W4206491577
FWCI: 0.072
Cites: 18
Cited by: 2
Related to: 10