2026-04-17T04:24:55Zhttps://recercat.cat/oai/request

oai:recercat.cat:2072/4850192025-08-31T18:14:36Zcom_2072_98col_2072_378192

00925njm 22002777a 4500 dc Porti i Pujal, Marc author Solis, Alvaro author Calatayud, Alex author Nafría i Maqueda, Montserrat author Miranda, Enrique author 2025 Catastrophic failure sites, also referred to as breakdown spots, in Metal-Insulator-Semiconductor (MIS) and Metal-Insulator-Metal (MIM) structures are the consequence of the formation of conducting paths across the thin oxide film that separates the contact electrodes. When the energy released by the sudden occurrence of this kind of shorts is high enough, the events are clearly detected in the top area of the structure as a random spatial point pattern. As it was demonstrated in previous works, the distribution of failure sites obtained this way can be used to generate optically detectable cryptographic keys in the context of Physically Unclonable Functions (PUFs). In this paper, we pay special attention to the reliability of the associated fingerprints. Reliability is evaluated in terms of a number of features of the binarized images such as the rotation and translation of the observation window, resolution, illumination, noise conditions, and particularities of the used optical system. The obtained results demonstrate that the generated fingerprints meet the essential requirements of reliability, reaching values between ~90 and 99% in all the considered scenarios. By means of a simulated experiment, which closely resembles the practical application of the proposed method, we are able to assess how good the identification of the registered images is and therefore the feasibility of the considered approach. To complete the picture, the investigated PUFs are shown to be resilient to temperature and electrical stress attacks which makes them highly suitable for security applications. http://hdl.handle.net/2072/485019 Breakdown Cryptography Dielectric breakdown MIM MIS Reliability assessment of optical physical unclonable functions based on the spatial distribution of catastrophic failure sites in MIM structures