# US 2026/0202464 A1 · Use of electromagnetic signature to determine the type of integrated circuit

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0202464 A1
- Document ID: US20260202464A1
- Status: Published application
- Kind: A1
- Application number: 19/557,910
- Filed: 2026-03-05
- Published: 2026-07-16
- Earliest claimed: 2021-06-02
- Assignees: Ohio State Innovation Foundation
- Applicants: Ohio State Innovation Foundation
- Inventors: Robert Lee; Aditya Nechiyil; Gregg Chapman
- Web page: https://patentgrep.com/patent/US20260202464A1

## CPC

- G01R 31/2839: {using signal generators, power supplies or circuit analysers (G01R31/2879 takes precedence; multimeters G01R15/12, network analysers G01R27/28)}
- G01R 31/2844: {using test interfaces, e.g. adapters, test boxes, switches, PIN drivers (G01R31/2889 takes precedence)}
- G01R 29/0814: {Field measurements related to measuring influence on or from apparatus, components or humans (EMC, EMI and similar testing in general G01R31/001), e.g. in ESD, EMI, EMC, EMP testing, measuring radiation leakage; detecting presence of micro- or radiowave emitters; dosimetry; testing shielding; measurements related to lightning}
- G01R 31/311: of integrated circuits {(G01R31/31728 takes precedence)}
- G01R 23/00: Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12)

## Abstract

Systems and methods for detecting anomalous integrated circuits (ICs) are provided. An electrical signature for each of multiple reference ICs are created using a resonator. Each IC may be associated with a different model or manufacturer. Later, when an IC that is suspected of being anomalous is received, its signature is generated and compared with a previously generated signature for an IC of the same make and manufacturer. When the signatures differ by more than a threshold amount, then the IC may be determined to be anomalous. In some embodiments, the generated signatures may be used to train a model that can determine whether an IC is anomalous based on its generated signature or can identify the make and manufacturer of an IC based on its generated signature.

## Claims

### Claim 1

A method comprising:
receiving a plurality of integrated circuits by a computing device, wherein each integrated circuit is associated with an identifier;
for each integrated circuit of the plurality of integrated circuits, generating a signature for the integrated circuit by the computing device;
generating a model using the signature generated for each integrated circuit by the computing device;
receiving a signature for a first integrated circuit not in the plurality of integrated circuits by the computing device; and
using the model and the received signature, predicting the identifier associated with the first circuit by the computing device.

### Claim 2 (depends on 1)

The method of claim 1, further comprising determining that the predicted identifier does not match an identifier associated with the first integrated circuit, and in response, determining that the first integrated circuit is an anomalous integrated circuit.

### Claim 3 (depends on 2)

The method of claim 2, wherein the analogous integrated circuit is one or more of a defective integrated circuit or a counterfeit integrated circuit.

### Claim 4 (depends on 1)

The method of claim 1, wherein the model is one or more of a random forest classifier, a k-nearest neighbor classifier, or a support vector machine.

### Claim 5 (depends on 1)

The method of claim 1, wherein generating the signature for the integrated circuit comprises generating the signature for the integrated circuit using a cavity resonator.

### Claim 6 (depends on 5)

The method of claim 5, wherein generating the signature for the integrated circuit using the cavity resonator comprises:
placing the integrated circuit in the cavity resonator at a first predetermined position;
providing an electromagnetic wave through the cavity resonator;
while the integrated circuit is in the first predetermined position, measuring a first resonate frequency in the cavity resonator; and
generating the signature based on the measured first resonate frequency.

### Claim 7 (depends on 6)

The method of claim 6, wherein generating the signature for the integrated circuit using the cavity resonator further comprises:
placing the integrated circuit in the cavity resonator at a second predetermined position;
while the integrated circuit is in the second predetermined position, measuring a second resonate frequency in the cavity resonator; and
generating the signature based on the measured first resonate frequency and the second resonate frequency.

### Claim 8 (depends on 6)

The method of claim 6, further comprising:
while the first integrated circuit is in the first predetermined position, measuring a return loss value of the integrated circuit and
generating the signature based on the measured first resonate frequency and the return loss value.

### Claim 9 (depends on 1)

The method of claim 1, wherein generating the signature for the integrated circuit comprises:
for each of a plurality of modes of a cavity resonator, measuring a resonant frequency value and a corresponding return loss value while the integrated circuit is positioned within the cavity resonator; and
generating the signature based on the plurality of measured resonant frequency values and return loss values.

### Claim 10 (depends on 1)

The method of claim 1, further comprising:
placing each integrated circuit of the plurality of integrated circuits in the cavity resonator at a plurality of predetermined positions and, for each position, measuring at least one resonant frequency value and at least one return loss value in the cavity resonator; and
generating the signature for the integrated circuit based on measurements obtained at the plurality of predetermined positions.

### Claim 11 (depends on 1)

The method of claim 1, wherein generating the model using the signature generated for each integrated circuit comprises:
clustering the signatures into a plurality of clusters and associating each cluster with at least one identifier corresponding to one or more of a type, model, or manufacturer of the integrated circuits.

### Claim 12 (depends on 1)

The method of claim 1, wherein generating the model using the signature generated for each integrated circuit comprises:
generating training data including, for each of a plurality of circuit types, differences between signatures of non-anomalous circuits labeled as non-anomalous and differences between signatures of anomalous and non-anomalous circuits labeled as anomalous; and
training the model using the labeled differences.

### Claim 13 (depends on 1)

The method of claim 1, wherein predicting the identifier associated with the first circuit comprises:
generating, by the model, a confidence value associated with the predicted identifier and outputting both the predicted identifier and the confidence value.

### Claim 14 (depends on 1)

The method of claim 1, further comprising:
when the confidence value associated with the predicted identifier is less than a threshold confidence value, determining that the first integrated circuit is possibly anomalous.

### Claim 15 (depends on 1)

The method of claim 1, wherein predicting the identifier associated with the first circuit comprises:
determining that the received signature corresponds to a signature that is not associated with any of a plurality of clusters of signatures generated from the plurality of integrated circuits and;
in response, determining that the first integrated circuit is anomalous.

### Claim 16 (depends on 1)

The method of claim 1, wherein the model is configured to output, based on the received signature, a probability distribution over a plurality of candidate identifiers, and wherein predicting the identifier associated with the first circuit comprises selecting, from the probability distribution, the identifier having a highest probability.

### Claim 17 (depends on 1)

The method of claim 1, wherein generating the signature for the integrated circuit comprises:
providing an electromagnetic wave through the cavity resonator over a frequency sweep that covers at least a fundamental transverse electric mode and one higher order transverse electric mode of the cavity resonator; and
generating the signature based on shifts in resonant frequencies and changes in return loss values for the fundamental transverse electric mode and the higher order transverse electric mode caused by insertion of the integrated circuit into the cavity resonator.
