Detection efficiency and noise in semi-device independent randomness extraction protocol

Abstract

In this paper, we analyze several critical issues in semi-device independent quantum information processing protocol. In practical experimental realization randomness generation in that scenario is possible only if the efficiency of the detectors used is above a certain threshold. Our analysis shows that the critical detection efficiency is 0.7071 in the symmetric setup, while in the asymmetric setup if one of the bases has perfect critical detection efficiency then the other one can be arbitrarily close to 0. We also analyze the semi-device independent random number generation efficiency based on different averages of guessing probability. To generate more randomness, the proper averaging method should be applied. Its choice depends on the value of a certain dimension witness. More importantly, the general analytical relationship between the maximal average guessing probability and dimension witness is given.

References in corpus (8)

• Device-independent quantum key distribution secure against collective attacks
• Device-independent tests of classical and quantum dimensions
• Semi-device-independent security of one-way quantum key distribution
• Mismatched-basis statistics enable quantum key distribution with uncharacterized qubit sources
• Quantum key distribution based on quantum dimension and independent devices
• Properties of dimension witnesses and their semi-definite programming relaxations
• Relation between semi- and fully-device-independent protocols
• Decoy state measurement-device-independent quantum key distribution based on the Clauser-Horne-Shimony-Holt inequality