Technische Universität Darmstadt

Bachelor Thesis: Security Evaluation of Physically Unclonable Functions

Online seit
16.01.2019 - 16:10
Job-ID
DA-2019-01-16-228772
Jobtyp

Beschreibung

Background

Since their invention, Physically Unclonable Functions have been introduced as central building blocks in security architectures and cryptographic protocols. A PUF is a noisy function embedded in an electronic device that generates a device-specific output to a given input. As their name implies PUFs were built with the goal of being unclonable and not emulatable. Researchers showed that machine learning algorithms can be used to learn models of PUFs architectures and that current PUF architectures do not reach that goal yet. New PUF architectures promise a better protection against the described modeling attacks. However, their resistance has not thoroughly been tested in research yet.

Task

Implement and calibrate recent PUF architectures based on their description in the literature. Implemented PUFs shall be simulated to collect their outputs. The final step is to evaluate their resilience to several Machine Learning algorithms, which are already provided. 

Anforderungsprofil

  • Background/Experience in Analog/mixed-signal design
  • Basic background in Cadence tools (Virtuoso & Spectre) or Synopsys
  • Familiar with verilog-a coding (can modify & write code of device’s model from description and equations )
  • Familiar with programming in Python / C++

If you want to become part of this interesting project and participate in state-of-the-art research, please send your application (CV, certificate(s) and any supporting documents) via email to Mrs. Shaza Zeitouni shaza.zeitouni@trust.tu-darmstadt.de

Art der Beschäftigung

nach Vereinbarung

Kontaktdaten

Ansprechpartner
Frau Mandy Herberg
Einsatzort
Mornewegstraße
30
64293 Darmstadt
Deutschland
Telefon
+49 6151 1625344
E-Mail

Weitere Informationen

Institut / Einrichtung
Systemsicherheit
Zeitraum der Beschäftigung
nach Vereinbarung
Bewerbungsfristende
Sonntag, 31. März 2019 - 23:59