Counterfeit IC Detection Using RF Excited Signals and AI-assisted Classification

Published

July 9, 2025

Author(s)

Jack Chuang, Chiehping Lai, David Griffith

Abstract

Counterfeit semiconductor devices are a major economic and security threat that can cause losses in multiple economic sectors. This problem's impact on national security is serious and growing because every part of the country, such as the military, utilities, businesses, and individuals, relies on complex and highly integrated circuits and systems. In 2019, the Organization for Economic Co-operation and Development (OECD) estimated that the global trade of pirated IC products is $500 billion [1]. Systems that detect counterfeit ICs, such as those that use X-ray microscopy, are challenging to deploy [2]. Using RF signatures to detect counterfeit IC detection requires precise setups, advanced signal processing techniques, EM modeling, and a wide frequency range of operation (MHz to mmWave). NIST is developing a counterfeit IC detection system that aims to be low-cost, allow rapid deployment, and have a high interception rate. This approach uses measurement and analysis tools based on excited and reflected RF signals from ICs under test that contain non-linear and non-deterministic signal behaviors.

Proceedings Title

The Symposium on Counterfeit Parts and Materials

Conference Dates

June 25-27, 2024

Conference Location

College Park, MD, US

Pub Type

Conferences

Download Paper

Local Download

Keywords

Counterfeit ICs, RF, Sensing, AI, Machine Learning

Citation

Chuang, J. , Lai, C. and Griffith, D. (2025), Counterfeit IC Detection Using RF Excited Signals and AI-assisted Classification, The Symposium on Counterfeit Parts and Materials, College Park, MD, US, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=957867 (Accessed July 12, 2025)