This course was developed as a MEST Micro-Certificate. MEST Micro-Certificates offer short, self-paced learning opportunities through focused training modules and webinars on specialized technical topics. Participants who successfully complete the program requirements earn a certificate of completion.
These programs are designed for U.S. citizens, including Department of War and other government personnel, government-affiliated employees, industry professionals, college students, and faculty.
Physical Unclonable Functions (PUFs) leverage small differences in the behavior of identically designed devices to generate device-specific keys, bitstrings and random numbers. A PUF architecture defines a source of entropy and an algorithm that measures and processes random variations which occur in the entropy source into a digital sequence of 0’s and 1’s. In this course, we will investigate several PUF architectures, and discuss the impact of adverse environmental conditions on the statistical quality and reliability of the generated bitstrings. A set of laboratory exercises is included to give trainees experience with designing PUFs on field programmable gate arrays.
PUFs are designed to measure and digitize electrical signal variations that occur in the device that occur because of the non-zero tolerance associated with the manufacturing process. These variations introduce small differences in identically designed devices, i.e., the delay along paths implemented with logic gates is different in each copy of the device. An important goal when designing a PUF is to select a circuit parameter that exhibits random behavior across devices, and to avoid circuit parameters that possess systematic bias. Systematic bias adversely impacts the statistical quality of the PUF-generated bitstrings because different copies of the device can produce similar bitstrings. We investigate several PUF architectures in this course, that incrementally improve on the ability of the PUF to produce high quality bitstrings. Three PUF architectures are discussed including:
Unit 1: Ring Oscillator (RO) PUF: An identically designed circuit structure that leverages delay variations as a source of entropy. The RO PUF is constructed as a localized circuit structure which creates a cyclic ring of gates which propagates a signal which behaves like a clock signal but whose frequency of oscillation is dependent on random variations that occur within the inverter gates defining the ring.
Unit 2: Shift-Register (SR) PUF: A hybrid architecture that includes some elements that are identically designed and others that are not identically designed. Similar to the RO PUF, the SR PUF propagates a signal around a loop, and variations in the frequency of oscillation defines the source of entropy.
Unit 3: Shift-Register, Reconvergent-Fanout (SiRF) PUF: A non-identically designed test structure that leverages variations that occur in path delays through an engineering netlist of logic gates. Unlike the RO and SR PUF, the delay variations are measured using a time-to-digital converter, and a sequence of post-processing operations are used to improve the statistical quality of the bitstrings.
Unit 4: RO PUF laboratory: A set of two modules that walk trainees through the process of designing a ring oscillator PUF on an FPGA.
Unit 5: Supplemental Material: A presentation on the types of embedded instrumentation that are useful for PUF architectures is also provided. In particular, PUFs that leverage resistance and delay variations as a source of entropy can utilize a voltage-to-digital converter (VDC) and a time-to-digital converter (TDC) to improve statistical metrics including the uniqueness and reliability of the PUF.
Format: Online, asynchronous
Platform: nanoHUB Learning Management System
Estimated Time: 10-12 hours
Level: Advanced Undergraduate/Graduate Student
Prerequisites:
A background in Very Large Scale Integration (VLSI) and Field Programmable Gate Arrays (FPGAs) is desirable but not required.
MEST Security Primitives I: Introduction to Physically Unclonable Functions
Instructor: Professor Jim Plusquellic, University of New Mexico
Course Objectives:
Compare PUF architectures using their underlying circuit variations
Design a Ring Oscillator PUF on an FPGA platform
Explain the role of post-processing in enhancing PUF statistical metrics
Apply voltage-to-digital and time-to-digital converters to PUF designs
Request Course Enrollment
To enroll in this course, you must have a nanoHUB account. Create one here: https://nanohub.org/register/
Request membership in the MEST group in nanoHUB at https://nanohub.org/groups/mest