Self-Adaption of Quantum Key Distribution Devices to Changing Working Conditions

Quantum key distribution (QKD) draws security from the ability to detect eavesdroppers upon the incident of “unnaturally high” quantum bit error rates (QBER) during the protocol. This presumes that a working QKD device implementation needs to be calibrated with its own individual and characteristic channel noise that is observed in any case, especially when there is no eavesdropper. This natural noise level influences the idle QBER, and is in turn dependent on the device’s working conditions. This work describes a statistical approach to construct a model that can be used to continuously compute the normal QBER, based on current (and changing) working conditions. The model is thus meant as a tool to refine the theoretical QBER threshold specific to the given QKD protocol, based on empirical data obtained under the given environmental conditions. More importantly, it allows the QKD-device to self-recalibrate under changing working conditions.