A Quantum Steganography Method Based on Directional Transformation of a Perfect Code Generator Matrix

Current quantum steganography techniques are primarily based on ground state encoding or probability amplitude embedding mechanisms. However, secret information remains confined to classical bit (0/1) formats, which makes it difficult to meet the processing demands of arbitrary superposition states α|0〉 + β |1〉 in fully quantum scenarios. This paper proposes a method that constructs codeword variants through directional transformation of a perfect code generator matrix. These variants exhibit statistical properties that closely resemble those of channel noise parameters under the stabilizer measurement framework, thereby ensuring that eavesdroppers cannot perceive the existence of steganography or decode it effectively. The proposed method offers the following advantages: (1) it achieves the unification of quantum state carriers and secret information, supporting direct steganography of arbitrary superposition states (see Fig. 1); (2) it adopts a Latin square balancing strategy to make the transformed perfect codes match the parameters of depolarizing channel noise, thus reducing the detectability of statistical analysis by an eavesdropper; (3) it uses error-correcting codes as carriers, which possess inherent noise resistance and are suitable for arbitrary Pauli errors. Numerical simulations under Pauli noise demonstrate that with gate noise intensities in (0.01,0.1), eavesdroppers’ quantum state fidelity remains bounded below 0.01 against arbitrary single-qubit Pauli errors, while legitimate users achieve fidelity > 0.93. This protocol proposes a robust quantum information hiding framework for quantum internet and IoT applications.