Hardware security assurance is a growing issue throughout the microelectronics and communications supply chains. There exist new opportunities to augment provenance-based approaches to hardware security with measurement-based techniques, whereby an individual device or component can be confidently identified or classified based on electrical measurements alone. This approach can provide additional information on a device’s security profile, potentially detecting counterfeit or unauthorized modifications at the component level.

 

We are developing detailed linear and nonlinear electromagnetic measurements of circuits and circuit elements at microwave and mm-wave frequencies with the goal of identifying or classifying an individual circuit element as trusted or suspicious. Intrinsic variations in the manufacturing or packaging processes can give rise to subtle differences in the detailed electromagnetic response that are the basis for classification and identification. A key element is the development of accurate and reproducible measurements of linear response and nonlinear intermodulation distortion and applying machine-learning techniques to classify devices as trusted or untrusted. Detailed multiphyics modeling approaches are also applied to help validate the machine learning algorithms developed to classify device linear and nonlinear response.

 

Reference: J.C. Booth et al., “5G Hardware Supply Chain Security Through Physical Measurements,” National Institute of Standards and Technology, Gaithersburg, MD, NIST SP 1278, 2022. doi: 10.6028/NIST.SP.1278.

Advanced Packaging; Electromagnetics; Hardware Security; Microelectronics; Microwave; mm-Wave; Nonlinear Measurements;

citizenship

Open to U.S. citizens

level

Open to Postdoctoral applicants