Title: Designing an SRAM PUF-based secret extractor for resource-constrained devices

Authors: Mario Barbareschi; Pierpaolo Bagnasco; Domenico Amelino; Antonino Mazzeo

Addresses: Department of Electrical Engineering and Information Technologies, University of Naples Federico II, Naples, NA, Italy ' Department of Electrical Engineering and Information Technologies, University of Naples Federico II, Naples, NA, Italy ' Department of Electrical Engineering and Information Technologies, University of Naples Federico II, Naples, NA, Italy ' Department of Electrical Engineering and Information Technologies, University of Naples Federico II, Naples, NA, Italy

Abstract: Extracting binary secrets from the physical imperfections, which affect every manufactured silicon device, guarantees attractive properties like randomness and uniqueness, physical unclonability and tamper evidence. Silicon physically unclonable functions (PUFs) were introduced aiming at providing keys from physically imprinted secrecy, instead of involving non-volatile memory as storage. However, the implementation of such a promising solution has to cope with unstable values provided by PUFs, requesting post-processing techniques to recover the secret from a noisy version. One of the most interesting PUF architecture exploits the pattern generated by static RAM cells when they are powered up, because many circuits embed them as non-volatile memory technology. Nevertheless, owing to the overhead introduced by the secret recovering, even if the SRAM is available to be used as PUF on a device, the design of secret extraction architecture must take into account the availability of requested computational resources. In this paper, we review one of the main techniques which accomplish the secret extraction from SRAM PUFs, namely the fuzzy extraction algorithm, detailing involved tools and design parameters for an implementation on a commercial device. For this purpose, we target the STM32F4 microcontroller family, tailoring every design parameter for the SRAM embedded in it, demonstrating the feasibility of the approach with an application based on the remote secure software update.

Keywords: static random access memory; SRAM; microcontroller; physically unclonable functions; PUFs; fuzzy extractor.

DOI: 10.1504/IJES.2017.086136

International Journal of Embedded Systems, 2017 Vol.9 No.4, pp.353 - 364

Received: 09 Dec 2015
Accepted: 25 Sep 2016
Published online: 27 Aug 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article