Ruhrmair et al. proposed a method to model PUF by using ES- and LR-based Machine Learning algorithms [32] , where the attackers may guess the internal linear delay vector after capturing many CRPs, so the simulated PUF is realized via software modeling [3] . Majzoobi et al. proposed the authentication verification model and used a PUF model for system security [25] in the following way: as the trusted manufacturer can access the PUF structure in the prover through the I/O pins of the chip and retrieve enough CRPs to train the PUF model, it is successfully built with the I/O pin permanently destroyed, and no longer can access the PUF via the I/O pin. Next, the trusted manufacturer sends the PUF model to the verifier, which stores a PUF model for each prover.