International Association
for Cryptologic Research

Quantum Readout PUFs (QR-PUFs) have been proposed as

a technique for remote authentication of objects. The security

is based on basic quantum information theoretic principles

and the assumption that the adversary cannot efficiently implement arbitrary unitary transformations.

We analyze the security of QR-PUF schemes in the case where

each challenge consists of precisely $n$ quanta and the dimension $K$ of the Hilbert space is larger than $n^2$.

We consider a class of attacks where the adversary first tries to learn as much as possible about the challenge and then bases his response on his estimate of the challenge.

For this class of attacks we derive an upper bound on the adversary\'s success probability as a function of $K$ and~$n$.