Noise Resilience of Quantum Key Distribution Protocols Secured Against Independent Attacks With One-Way Communication

Quantum Key Distribution (QKD) is a cornerstone of quantum communication, theoretically offering information-theoretic security. However, practical implementations are plagued by physical noise, which degrades the secret key rate and can potentially be exploited by an eavesdropper. While the Quantum Bit Error Rate (QBER) is the standard metric for assessing these systems, this paper argues that QBER is an insufficient indicator of true physical noise resilience.

The authors introduce a new noise-based metric, denoted by ε, which quantifies the cumulative deviation of a practical QKD implementation from an ideal, noiseless system. This metric accounts for imperfections across state preparation, transmission, and measurement. By defining a noise threshold Υ—the maximum physical noise level allowing for secure key distillation—the researchers use semidefinite programming (SDP) to evaluate and compare the noise tolerance of several prominent protocols, including BB84, B92, E91, and the six-state protocol.

The study proves that the fundamental noise threshold for QKD protocols under independent attacks is 0.25. The six-state protocol is identified as the most robust, reaching this theoretical maximum. In contrast, other protocols like BB84, E91, and B92 exhibit significantly lower thresholds. Furthermore, the authors construct a family of protocols that can tolerate high empirical QBER while possessing low physical noise resilience, highlighting that empirical error rates are not always reliable proxies for the underlying physical robustness of a protocol.

This work provides a more rigorous framework for evaluating QKD security than the conventional QBER metric. By decoupling empirical bit errors from physical noise, the researchers offer a clearer path for assessing the security of practical quantum hardware. This framework helps developers identify which protocols are truly robust against environmental and adversarial noise, rather than those that merely appear resilient due to specific measurement configurations.