“…To resist machine learning attacks and detect any possible invasive attacks, different research groups have implemented further countermeasures. These involve integrating additional software functionalities, such as hashing, non-linear functions, and ephemeral CRP tables, or incorporating hardware solutions like XOR gates, CMOS, and other logic circuits [59], [60], [61], [62], [63], [64], [65]. For instance, [66], [67], [68] developed an obfuscated challenge-response protocol that utilizes a PUF chip, a random number generator (RNG), and a control block to safeguard against machine learning attacks, circumventing the need for traditional cryptographic techniques.…”