RadioFrequency Antenna Configuration, Evaluation, and Optimization
Univ. of California, San Diego , CA, USA
Oracle Principal Investigator
Guang Wang, Principal Member of Technical Staff
Kenny Gross, Architect
Counterfeit electronics has become a $250B per year problem across all electronics industries, including military, medical, and transportation electronics, and all forms of consumer electronics. US NIST has documented that the international distribution of counterfeit electronic components is 900% more profitable than the international distribution of cocain....but harder to detect.
This collaborative research project evaluates and optimizes a novel Oracle technology for passive detection of counterfeit electronics where EMI emissions for operating assets are detected with low-cost hand-held or magnetic-mounted units where the EMI emissions are analyzed with an advanced Machine Learning algorithm called the Multivariate State Estimation Technique (MSET). It has been demonstrated that the patterns of emission from any asset comprising a collection of internal electronic or electromechanical components emits a complex frequency pattern called an "EMI Fingerprint", which is as complex and unique for an asset as a snowflake. By training on a "Golden System EMIF" for an asset certified to have only authentic components inside, the EMIF technique can now scan any number of identical assets in the field, or in the supply chain, for passive detection of internal counterfeit components.