Apparatus and method for integrated circuit forensics

Patent metadata

Field	Value
Publication number	US-9885745-B2
Application number	US-201614996554-A
Country	US
Kind code	B2
Filing date	Jan 15, 2016
Priority date	Jun 24, 2013
Publication date	Feb 6, 2018
Grant date	Feb 6, 2018

Abstract

A test system including an embodiment having a sensor array adapted to test one or more devices under test in learning modes as well as evaluation modes. An exemplary test system can collect a variety of test data as a part of a machine learning system associated with known-good samples. Data collected by the machine learning system can be used to calculate probabilities that devices under test in an evaluation mode meet a condition of interest based on multiple testing and sensor modalities. Learning phases or modes can be switched on before, during, or after evaluation mode sequencing to improve or adjust machine learning system capabilities to determine probabilities associated with different types of conditions of interest. Multiple permutations of probabilities can collectively be used to determine an overall probability of a condition of interest which has a variety of attributes.

First claim

The invention claimed is:

1. An apparatus for testing including measuring electromagnetic emissions and determining defective components comprising:
a sensor array comprising a plurality of sensors adapted to be moveable;
a signal analysis section comprising a section comprising a time domain and signal domain signal analysis signal section;
a device under test (DUT) holder adapted to hold and position a first and second DUT relative to the sensor array;
a DUT stress application section adapted to generate one or more DUT stress conditions to said first DUT which correlate with life cycle age or a life cycle reduction event associated with said DUT, where said DUT stress condition associated with life cycle age is applied by an escalating series of burn-in or heating conditions applied to said first DUT, said DUT life cycle reduction event comprises electrostatic discharge application to at least one section of said first DUT;
a control mechanism adapted to independently position elements of said sensor array relative to areas of interest on said first and second DUT based on a first position input;
a DUT control section comprising a machine readable storage medium adapted to receive and store a plurality of machine readable instructions operable to control said apparatus and said first and second DUTs, said machine readable storage medium further comprises a first plurality of machine readable instructions adapted to operate said control mechanism in order to acquire and store a plurality of first sensor array signature data associated with said sensor array outputs from said first DUT based on a first plurality of test signal control inputs to said first DUT and said first position input while one or more said DUT stress conditions are applied to said first DUT, wherein each of said plurality of said first sensor array signature data is respectively associated with said one or more DUT stress conditions;
wherein said DUT control section further comprises a second plurality of machine readable instructions stored on said machine readable storage medium adapted to stimulate said second DUT when said second DUT is placed in said DUT holder with said first plurality of test signal control inputs, said DUT control section further comprises a third plurality of machine readable instructions stored on said machine readable storage medium adapted to acquire a plurality of second sensor array signature data associated with said sensor array outputs from said second DUT based on said first plurality of test signal control inputs to said second DUT and said first position input;
wherein said DUT control section further comprises a fourth plurality of machine readable instructions stored on said machine readable storage medium adapted to match said first and second sensor array signature data associated respectively with said first and second DUT, wherein a substantial match of said signature data indicates a first condition associated with said second DUT and a non-match indicates a second condition associated with said second DUT;
an input and output section adapted to interact with said DUT control section, said input and output section comprising a user interface including a graphical user interface adapted to display an indication of said first or second condition associated with said second DUT.

2. An apparatus as in claim 1 , wherein said first condition comprises a DUT acceptable data indicator and said second condition comprises a DUT unacceptable data indicator.

3. An apparatus as in claim 1 , wherein said fourth plurality of machine readable instructions comprises a decision engine and rule base comprising a plurality of data sets associated with a plurality of conditions comprising said first and second conditions, said decision engine and said rule base are operable for determining a probability that said second DUT is associated with one or more of said plurality of conditions, wherein said first condition comprises an authorized or meets specification condition and said second condition comprises an unauthorized, does not meet specification, or is defective condition.

4. A testing system including a system for measuring electromagnetic emissions and determining defective components comprising:
a sensor system means adapted to test one or more devices under test (DUT) in learning as well as evaluation modes, said sensor system means is adapted to collect a variety of test data as a part of a machine learning system associated with a known-good said one or more DUT samples subjected to simulated or generated condition of interests comprising stimulation of sections of said known-good DUT and application of one or more stress events to said known-good DUT sample, said data collected by the machine learning system is operable to calculate probabilities data that an unknown-good DUT in an evaluation mode substantially matches a condition of interest comprising a DUT acceptance data or a DUT rejection data; and
an input section and an output section adapted to output said probabilities data associated with said unknown-good DUT;
wherein learning modes are switched on before, during, or after evaluation mode sequencing to improve or adjust machine learning system capabilities to determine said probabilities associated with different types of said conditions of interest wherein said system are adapted to determine multiple permutations of said probabilities that are collectively used to determine an overall probability of one or more said conditions of interest which has a variety of attributes associated with one or more said DUT acceptance data and DUT rejection data.

5. A system as in claim 4 , wherein said learning and evaluation system include an artificial intelligence system comprising a neural network or a decision engine comprising a plurality of rules associated with generating said probabilities data.

6. A system as in claim 4 , wherein said learning and evaluation system are adapted for determining a probability that DUT comprising a microelectronic device is unauthorized, does not meet specification(s), or is defective.

7. A testing system comprising:
a plurality of device under test (DUT) measurement and data collection/input sections comprising electromagnetic (EM) spectrum sensors and data collection sections adapted to apply a plurality of stimulation inputs to a plurality of points on a known-good and an unknown-good DUT and respectively sense a first and second plurality of test data from said known-good and unknown good DUTs, wherein said plurality of stimulation inputs applied to said known-good DUT is applied during synchronized application of one or more life cycle simulation stress environments are applied to said known-good DUT comprising one or more of a group comprising electrostatic stress and ageing events comprising burn-in or heating of said known-good DUT associated with operational or functional capability conditions of said known-good DUT during specific points of said known-good DUT's service life;
one or more decision engine sections respectively associated with said plurality of DUT measurement and data collection/input sections, said one or more decision engine sections comprising a neural network, image recognition, statistical correlation section, and decision tree section, said one or more decision engines adapted to operate said DUT in a learning mode and an evaluating mode, said one or more decision engines are further adapted to receive said first and second test data from said collection/input modes at a first and second stage operable to enable said learning mode associated with said operational or functional capability conditions of said known-good DUT during its service life;
a control section operable to input said first and second test

Assignees

• Us Navy

Inventors

• Hamilton Brett J

Classifications

• G01R1/07

  Non contact-making probes · CPC title

• G01R31/2872

  related to electrical or environmental aspects, e.g. temperature, humidity, vibration, nuclear radiation · CPC title

• G06N3/02 Primary

  Neural networks · CPC title

• G06F11/2263

  using neural networks · CPC title

• G01R31/002

  where the device under test is an electronic circuit · CPC title