Authors:

Kevin Yu

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Counterfeit Integrated Circuits are repackaged, remarked, or cloned versions of current chips. They often do not meet the specifications of authentic chips, and using them in place of genuine parts puts systems at risk of malfunction, damage, and malware/spyware. In recent years, counterfeiters have improved their processes to evade existing forms of counterfeit detection. A procedure to quantify the efficacy of novel nondestructive screening methods, as well as combine their results into a multimodal classification model, remains desired. Our group explores several nondestructive analyses for verifying device authenticity using Spectrophotometry, X-Ray Spectroscopy, and Profilometry measurements. Using 1 authentic and 4 potential counterfeit sources of Raspberry Pi chips, we collected spectral, elemental, or photographic data from an Analytical Balance, X-ray Fluorescence (XRF), X-ray Photoelectron Spectroscopy (XPS), UV-Vis-NearIR Spectroscopy, Laser Induced Beam Spectroscopy (LIBS), Gas Chromatography (GC), Optical Microscopy, UV Fluorescence Microscopy (UVF), Contact Profilometry, and Laser Profilometry. We employ statistical analysis and machine learning (ML) to distinguish features of the counterfeit groups. For each of the 10 methods, we estimate the cost and time of measurement, cost and time of training data collection, cost and time of ML training, and confidence in distinguishing counterfeits. We recommend the top three methods based on our decision matrix, and note that anyone can change the weights to create a priority list to align with their constraints and values. We demonstrate that the results of multi-modal measurements can be combined to greatly increase the confidence in device authenticity. † Presenting Undergrad Author; ‡ Contributing Undergrad Author; * Undergrad Acknowledgment Keywords: Nondestructive; Counterfeit; Spectroscopy; Profilometry; Multimodal

Source:

Purdue University / 2025

Topics:

No topics listed

Co-authors:

Kevin Yu