Systems and methods for measuring unique microelectronic electromagnetic signatures

Patent metadata

Field	Value
Publication number	US-12416593-B1
Application number	US-202117329370-A
Country	US
Kind code	B1
Filing date	May 25, 2021
Priority date	Jun 1, 2017
Publication date	Sep 16, 2025
Grant date	Sep 16, 2025

Abstract

Systems and methods for measuring unique microelectronic electromagnetic signatures are provided. A method includes injecting a nondestructive signal as input into a port of an object. The method may further include receiving as output from a signal path within the object a unique frequency dependent complex spectrum comprising a reflection spectrum or a transmission spectrum. The method may also include generating a unique object signature based upon the port and the received spectrum. The method may still further include differentiating the object from a different object based upon a comparison of the unique object signature of each.

First claim

The invention claimed is:

1. A method for non-destructively testing an electronic suspect object to determine whether the suspect object is counterfeit or authentic, the method comprising:
injecting, via a signal injector, a known signal into one or more ports of a suspect object while the suspect object is in a powered-off state, the known signal comprising a non-destructive radio frequency signal;
identifying a signal response comprising a received spectrum responsive to the injecting of the known signal into the one or more ports of the suspect object;
generating a unique suspect object signature of the suspect object based at least in part on the signal response;
obtaining, from a memory, information indicative of a unique authentic object signature,
wherein the unique authentic object signature is based at least in part on a signal response of an authentic object;
generating a dataset comprising a plurality of object signatures including the unique suspect object signature and the unique authentic object signature;
performing a principal component analysis on the dataset comprising the plurality of object signatures such that principal component data points of unique objects are grouped into distinct clusters;
utilizing a probability density function and a transform after the principal component analysis to determine a confidence interval; and
differentiating the suspect object from the authentic object based at least in part on the principal component analysis.

2. The method of claim 1 , wherein injecting of the known signal into the one or more ports of the suspect object comprises injecting through a direct and physical connection between the signal injector and the one or more ports of the suspect object.

3. The method of claim 2 , wherein the known signal is measured using passive radio frequency injection spectrometry for characterizing a unique microelectronic electromagnetic signature of the known signal.

4. The method of claim 1 , wherein injecting of the known signal into the suspect object generates a reflected wave, wherein the reflected wave is measured to determine a distance the known signal travels into the suspect object.

5. The method of claim 1 , wherein the information indicative of the unique authentic object signature includes at least one of a manufacturer, data code, usage wear, wafer, packing house, fabrication location, age, environmental effects, or manufacturer effects.

6. The method of claim 2 , wherein there is a physical link between variation sources and real-space circuit locations and electromagnetic frequencies of the suspect object.

7. The method of claim 1 , wherein the signal response is measured at suspect object ports other than at a point of injection.

8. The method of claim 1 , further comprising:
storing a dataset comprising a plurality of unique object signatures from a plurality of objects.

9. The method of claim 1 , wherein the received spectrum comprises a unique frequency dependent complex spectrum comprising a reflection spectrum or a transmission spectrum.

10. The method of claim 1 , wherein the unique suspect object signature is based further in part on the one or more ports of the suspect object.

11. A system comprising at least one data processor and at least one non-transitory memory storing computer executable instructions that when executed by the at least one data processor cause the system to carry out actions comprising:
injecting, via a signal injector, a known signal into a plurality of ports of a suspect object while the suspect object is in a powered-off state, the known signal comprising a non-destructive radio frequency signal;
identifying a signal response comprising a received spectrum responsive to the injecting of the known signal into the plurality of ports of the suspect object;
generating a unique suspect object signature of the suspect object based at least in part on the signal response and the plurality of ports of the suspect object;
obtaining, from a memory, information indicative of a unique authentic object signature,
wherein the unique authentic object signature is based at least in part on a signal response of an authentic object;
generating a dataset comprising a plurality of object signatures including the unique suspect object signature and the unique authentic object signature;
performing a principal component analysis on the dataset comprising the plurality of object signatures such that principal component data points of unique objects are grouped into distinct clusters;
utilizing a probability density function and a transform after the principal component analysis to determine a confidence interval; and
differentiating the suspect object from the authentic object based at least in part on the principal component analysis.

12. The system of claim 11 , wherein the known signal is injected into the plurality of ports of the suspect object while the suspect object is in the powered-off state.

13. The system of claim 12 , further comprising a plurality of suspect objects, wherein each suspect object of the plurality of suspect objects is injected with the known signal to generate a unique object signature for each suspect object.

14. The system of claim 11 , wherein the signal injector comprises a multi-port vector network analyzer system and each multi-port vector network analyzer of the multi-port vector network analyzer system is configured for collecting information indicative of the suspect object between any two ports of the multi-port vector network analyzer system.

15. One or more non-transitory, computer readable media storing computer-executable instructions that, when executed by a processor, perform a method of performing non-destructive testing on a suspect object by comparing the suspect object with a known object using unique object identifiers, the method comprising:
injecting, via a signal injector, a known signal into one or more ports of a suspect object while the suspect object is in a powered-off state, the known signal comprising a non-destructive radio frequency signal;
identifying a signal response comprising a received spectrum responsive to the injecting of the known signal into the one or more ports of the suspect object;
generating a unique suspect object signature of the suspect object based at least in part on the signal response;
obtaining, from a memory, information indicative of a unique authentic object signature,
wherein the unique authentic object signature is based at least in part on a signal response of an authentic object;
generating a dataset comprising a plurality of object signatures including the unique suspect object signature and the unique authentic object signature;
performing a principal component analysis on the dataset comprising the plurality of object signatures such that principal component data points of unique objects are grouped into distinct clusters;
utilizing a probability density function and a transform after the principal component analysis to determine a confidence interval; and
differentiating the suspect object from the authentic object based at least in part on the principal component analysis.

16. The media of claim 15 , the method further comprising generating a cluster, wherein similar objects will form the cluster, with an unknown object appearing at a point some distance away from a centroid or an outer bounds of the cluster.

17. The media of claim 16 , wherein an unknown object that is a distance past a predetermined threshold is deemed to have failed a counterfeit detection test and a dist

Assignees

• Applied Res Associates Inc

Inventors

• Eakins Christian
• Kent Thomas

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G01N27/026 Primary

  Dielectric impedance spectroscopy (electrochemical impedance spectroscopy for measuring corrosion G01N17/02) · CPC title

• G06N20/00 Primary

  Machine learning · CPC title