Quantum Audio LSB Steganography with Entanglement-assisted Modulation

Abstract

Abstract This paper introduces two novel quantum audio steganography methods that incorporate the logical properties of entangled qubits into the least significant qubit steganography (LSQb) approach. The first method, called entanglement-assisted probability adjustment LSB (EP-LSQb), optimizes the superposition probability of the lowest qubits before embedding the information. This addresses the ‘value-pairing’ problem and improves anti-statistical analysis performance and embedding efficiency. The second method, called quantum phase direct embedding LSB (PD-LSQb), extends traditional LSB steganography by directly embedding the secret speech into the least significant qubit. Quantum circuits are constructed for both methods to demonstrate their feasibility. Theoretical analysis and simulation experiments show that these methods exhibit good imperceptibility (\({\text{PSNR}}>50\)) and embedding efficiency. Particularly, EP-LSQb method enhances the theoretical limit of embedding efficiency when compared to its corresponding classical counterparts. Additionally, due to the inclusion of entangled qubit particles, our methods offer competitive content security and noise robustness, distinguishing them from classical LSB or general LSQb methods and highlighting the unique advantages of quantum approaches.