A physical unclonable function (PUF) is an embedded integrated circuit (IC) structure that is designed to leverage naturally occurring variations to produce a random bitstring. In this paper, we evaluate a PUF which leverages resistance variations which occur in transmission gates (TGs) of ICs. We also investigate a novel on-chip technique for converting the voltage drops produced by TGs into a digital code, i.e., a voltage-to-digital converter (VDC). The analysis is carried out on data measured from chips subjected to temperature variations over the range of -40 o C to +85 o C and voltage variations of +/-10% of the nominal supply voltage. The TG PUF and VDC produce high quality bitstrings that perform exceptionally well under statistical metrics including stability, randomness and uniqueness.

Keying material for encryption is stored as digital bitstrings in non-volatile memory on FPGAs and ASICs in current technologies. However, secrets stored this way are not secure against a determined adversary, who can use probing attacks to steal the secret. Physical unclonable functions (PUFs) have emerged as an alternative. PUFs leverage random manufacturing variations as the source of entropy for generating random bitstrings, and incorporate an on-chip infrastructure for measuring and digitizing the corresponding variations in key electrical parameters, such as delay or voltage. PUFs are designed to reproduce a bitstring on demand and therefore eliminate the need for on-chip storage. In this paper, we evaluate the randomness, uniqueness and stability characteristics of a PUF based on metal wire resistance variations in a set of 63 chips fabricated in a 90 nm technology. The stability of the PUF and an on-chip voltageto-digital converter are evaluated at 9 temperature-voltage corners. Keywords -Physical Unclonable Function, power grid, metal resistance variationsming distance (HD) is used to determine the uniqueness of the bistrings among the population of chips. Similarly, the NIST statistical test suite can be used to evaluate the randomness of the bistrings produced by each chip [1]. And intra-chip HD can be used to evaluate stability of the bitstrings, i.e., the ability of each chip to reproduce the same bitstring time-after-time, under varying temperature and voltage conditions.In this paper, we focus on determining the temperature and voltage (TV) stability of a PUF that is based on resistance variations which occur in the metal wires of the chip's power grid. A significant benefit of using metal structures is that "noise-related" variations, such as those introduced by TV variations, result in linear changes to the measured voltages. This linear scaling characteristic allows the relative magnitude of two voltages to remain consistent across changes in temperature and voltage, which, in turn, improves the stability of the PUF to bit-flips 1 , when compared, for example to PUFs which leverage transistor-based variations.In our experiments, we evaluate the power grid (PG) PUF at 9 TV corners, i.e., over all combinations of 3 temperatures; -40 o C, 25 o C and 85 o C, and 3 voltages; nominal and +/-10% of nominal. The evaluation is carried out on a set of chips fabricated in IBM's 90 nm, 9 metal layer bulk silicon process. The stability of the bitstrings is measured using intra-chip HD and 'probability of failure' techniques. Randomness and uniqueness are also evaluated using the NIST test suite and inter-chip HD methods. A bit-flip avoidance scheme is proposed and evaluated that reduces the probability of a failure to reproduce the bitstring to less than 1E-9. We also investigate an on-chip voltage-to-digital converter (VDC) for measuring voltage variations (which reflect resistance variations in the metal wires) and its stability across the 9 TV corners.

No abstract

No abstract