Thermal Scans for Detecting Hardware Trojans

Maxime Cozzi 1 Philippe Maurine 1 Jean-Marc Galliere 2
1 SmartIES - Smart Integrated Electronic Systems
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
2 TEST - TEST
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
Abstract : It is well known that companies have been outsourcing their IC production to countries where it is simply not possible to guarantee the integrity of final products. This relocation trend creates a need for methodologies and embedded design solutions to identify counterfeits but also to detect potential Hardware Trojans (HT). Hardware Trojans are tiny pieces of hardware that can be maliciously inserted in designs for several purposes ranging from denial of service, programmed obsolescence etc. They are usually stealthy and characterized by small area and power overheads. Their detection is thus a challenging task. Various solutions have been investigated to detect Hardware Trojans. We focus in this paper on the use of thermal near field scans to that aim. Therefore we first introduce and characterize a low cost, large bandwidth (20 kHz) thermal scanning system with the high detectivity required to detect small Hardware Trojans. Then, we experimentally demonstrate its efficiency on different test cases.
Keywords : Trojan detection Lock-in thermography Thermal mapping Thermal modeling
Document type :
Conference papers
Complete list of metadatas
https://hal-lirmm.ccsd.cnrs.fr/lirmm-01823444
Contributor : Philippe Maurine <>
Submitted on : Tuesday, June 26, 2018 - 9:24:56 AM
Last modification on : Wednesday, May 8, 2019 - 2:56:01 PM

Identifiers

Collections

LIRMM | SMARTIES | TEST | LIRMM_MIC | MIC | MIPS | UNIV-MONTPELLIER

Citation

Maxime Cozzi, Philippe Maurine, Jean-Marc Galliere. Thermal Scans for Detecting Hardware Trojans. COSADE: Constructive Side-Channel Analysis and Secure Design, Apr 2018, Singapour, Singapore. pp.117-132, ⟨10.1007/978-3-319-89641-0_7⟩. ⟨lirmm-01823444⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

129

Contact

support.ccsd.cnrs.fr
hal.support@ccsd.cnrs.fr
Logo CCSD