Home Computer Science
Table of Contents:
Notions on Silicon Physically Unclonable Functions
The opportunity of extracting physical characteristics from fabric-induced variability for integrated circuits (ICs) has been representing the most important breakthrough for the semiconductor security. Like human fingerprints, retina blood vessels and DNA, which are the main identification means in the biometrics field, physically imprinted random patterns were adopted for the identification of serial manufacturing products [36,42]. Such a technological innovation led to formally define the Physical Random Function , and then the physical(ly) unclonable function, or PUF, which are able to identify a silicon device by means of random manufacturing variability.
During last years, PUF has begun a hot topic in the field of hardware security and trust, in fact countless PUF circuits and architectures have been proposed in the literature. Actually, more and more proposed secure infrastructures and applications rely on the adoption of PUFs, specially because they guarantee extremely attractive properties, such as uniqueness, unclonability, anti-tamper, and intrinsically randomness, without requiring any modification to the classical photolithography manufacturing process.
The goal of this chapter is to collect most of concepts related to PUFs which have been divulged in scientific papers, trying to give a uniform view of formal notions and quality metrics. Furthermore, some pointers regarding the security and reliability issues are summed up, covering indispensable aspects to implement any PUF-based application
This chapter is structured as follows. Section 10.2 gives a formal characterization of PUFs and related terminology; and lists PUF properties, mostly covering what has been introduced in the literature. Moreover, it contains the related terminology, necessary to comprehend further notions and concepts. Section 10.3 concertizes
M. Barbareschi (B)
DIETI—Department of Electrical Engineering and Information
Technologies, University of Naples Federico II, Via Claudio, 21 - 80125 Naples, Italy
© Springer International Publishing Switzerland 2017 N. Sklavos et al. (eds.), Hardware Security and Trust, DOI 10.1007/978-3-319-44318-8_10
previously defined properties by means of quality parameters (such as uniqueness, Sect. 10.3.1, and reliability, Sect. 10.3.2), which are used to compare and analyze available PUF architectures. Section 10.4 classifies existing PUF architectures by means of the source of randomness that their exploit.