Quantum Reversible Audio Steganalysis Using Quantum Schmidt Decomposition and Quantum Support Vector Machine

Abstract

With the development of various methods of quantum steganography, the discovery of these secret communications has become a necessity. This paper presents the first quantum steganalysis approach that uses statistical features of the quantum frequency domain to detect the steganographic methods in quantum communication networks. In this method, by proposing different quantum circuits, statistical features such as quantum mean, quantum square mean root, quantum variance, and quantum standard deviation are extracted to increase the accuracy of distinguishing between stego and clean signals. In the proposed method, the Schmidt coefficient vector extracted from the Quantum Haar Wavelet Transform (QHWT) is used to calculate the statistical features of the audio signal so that the main and dominant coefficients below the frequency sub-bands of the audio signal can be identified and selected. The proposed method uses a Quantum Support Vector Machine (QSVM) to classify the extracted feature vectors. The higher accuracy of the proposed frequency-based steganalysis method indicates its higher efficiency in detecting stego and clean signals than the quantum steganalysis methods based on the time domain.