From key distribution to direct transmission: a comprehensive review of quantum image security via QKD and QSDC

Abstract

In the rapidly evolving landscape of cybersecurity, traditional cryptographic systems are increasingly vulnerable to attacks, including brute-force, side-channel, man-in-the-middle, replay, and ransomware attacks, highlight the limitations of classical encryption techniques. Quantum cryptography leverages the no-cloning theorem and the properties of quantum states to establish fundamentally secure communication protocols with intrinsic eavesdropping detection capabilities. This security framework provides information-theoretic protection beyond the mathematical assumptions underlying conventional cryptographic systems. Quantum image security has evolved into two major paradigms: Quantum Key Distribution (QKD) and Quantum Secure Direct Communication (QSDC). Although recent surveys have reviewed both approaches chronologically, they have not systematically analysed their security thresholds The objective of this paper proposes a three-axis taxonomy of QSDC protocols, classifying them by quantum resource type, physical transmission channel, and device trust model. Furthermore, it presents a comparative performance analysis of QKD employing a hyperchaotic cipher over QSDC channels across seven quantum image representations, including FRQI, NEQR, GQIR, and MCQI. The analysis shows that QKD-seeded schemes achieve efficient key distribution, whereas pixel-level security remains dependent on cipher complexity. In contrast, QSDC provides end-to-end security governed by quantum mechanical principles; hyperentangled carriers achieve an eavesdropping detection probability of 0.875 compared with 0.5 for conventional two-step protocols, although communication throughput remains a limiting factor. Based on these findings, this review outlines future research directions, including QSDC-specific quantum repeaters for continental-scale deployment, hyperentangled carriers supporting up to 12 bits per photon pair compatible with NEQR’s 8-bit encoding, and machine-learning-assisted management of hybrid fiber–free-space quantum communication networks.